tag:theconversation.com,2011:/us/topics/search-engines-6341/articlesSearch engines – The Conversation2023-12-05T22:42:16Ztag:theconversation.com,2011:article/2185172023-12-05T22:42:16Z2023-12-05T22:42:16ZWikipedia’s volunteer editors are fleeing online abuse. Here’s what that could mean for the internet (and you)<figure><img src="https://images.theconversation.com/files/562311/original/file-20231129-17-hg57m4.jpg?ixlib=rb-1.1.0&rect=44%2C11%2C7304%2C4120&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>We’re now sadly used to seeing toxic exchanges play out on social media platforms like X (formerly Twitter), Facebook and TikTok. </p>
<p>But Wikipedia is a reference work. How heated can people get over an encyclopedia? </p>
<p>Our <a href="https://academic.oup.com/pnasnexus/article-lookup/doi/10.1093/pnasnexus/pgad385">research</a>, published today, shows the answer is very heated. For example, one Wikipedia editor wrote to another:</p>
<blockquote>
<p>i will find u in real life and slit your throat.</p>
</blockquote>
<p>That’s a problem for many reasons, but chief among them is if Wikipedia goes down in a ball of toxic fire, it might take the rest of the internet’s information infrastructure with it. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/let-the-community-work-it-out-throwback-to-early-internet-days-could-fix-social-medias-crisis-of-legitimacy-213209">Let the community work it out: Throwback to early internet days could fix social media's crisis of legitimacy</a>
</strong>
</em>
</p>
<hr>
<h2>The internet’s favourite encyclopedia</h2>
<p>In some ways, Wikipedia is both an encyclopedia and a social media platform. </p>
<p>It’s <a href="https://www.semrush.com/website/top/">the fourth most popular website</a> on the internet, behind only such giants as Google, YouTube and Facebook. </p>
<p>Every day, <a href="https://stats.wikimedia.org/#/all-projects">millions of people worldwide</a> use it for quick fact-checks or in-depth research. </p>
<p>And what happens to Wikipedia matters beyond the platform itself because of its central role in online information infrastructure. </p>
<p>Google search relies heavily on Wikipedia and the quality of its search results would <a href="https://ojs.aaai.org/index.php/ICWSM/article/view/14883/14733">decrease substantially</a> if Wikipedia disappeared. </p>
<p>But it’s not just an increasingly authoritative source of knowledge. Even though we don’t always lump Wikipedia in with other social media platforms, it shares some common features. </p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1724890300441895152"}"></div></p>
<p>It relies on contributors to create the content that the public will view and it creates spaces for those contributors to interact. Wikipedia relies solely on the work of volunteers: no one is paid for writing or editing content. </p>
<p>Moreover, no one checks the credentials of editors — anyone can make a contribution. This arguably makes Wikipedia the most successful collaborative project in history. </p>
<p>However, the fact that Wikipedia is a collaborative platform also makes it vulnerable. </p>
<p>A <a href="https://meta.wikimedia.org/wiki/Research:Harassment_survey_2015">2015 survey</a> found 38% of surveyed Wikipedia users had experienced harassment on the platform.</p>
<p>What if the collaborative environment deteriorates, and its volunteer editors abandon the project? </p>
<p>What effect do toxic comments have on Wikipedia’s editors, content and community?</p>
<h2>Abusive comments lead to disengaging</h2>
<p>To answer this question, we started with Wikipedia’s “user’s talk pages”. A user’s talk page is a place where other editors can interact with the user. They can post messages, discuss personal topics, or extend discussions from an article’s talk page. </p>
<p>Every editor has a personal user’s talk page, and the majority of toxic comments made on the platform are on these pages. </p>
<p>We collected information on 57 million comments made on the user’s talk pages of 8.5 million editors across the six most active language editions of Wikipedia (English, German, French, Italian, Spanish and Russian) over a period of 20 years. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/students-are-told-not-to-use-wikipedia-for-research-but-its-a-trustworthy-source-168834">Students are told not to use Wikipedia for research. But it's a trustworthy source</a>
</strong>
</em>
</p>
<hr>
<p>We then used a <a href="https://perspectiveapi.com/">state-of-the-art machine learning algorithm</a> to identify toxic comments. The algorithm looked for attributes a human might consider toxic, like insults, threats, or identity attacks.</p>
<p>We compared the activity of editors before and after they received a toxic comment, as well as with a control group of similar editors who received a non-toxic rather than toxic comment. </p>
<p>We found receiving a single toxic comment could reduce an editor’s activity by 1.2 active days in the short term. Considering that 80,307 users on English Wikipedia alone have received at least one toxic comment, the cumulative impact could amount to 284 lost human-years. </p>
<p>Moreover, some users don’t just contribute less. They stop contributing altogether. </p>
<p>We found that the probability of leaving Wikipedia’s community of contributors increases after receiving a toxic comment, with new users being particularly vulnerable. New editors who receive toxic comments are nearly twice as likely to leave Wikipedia as would be expected otherwise. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="The wikipedia logo on a yellow office wall" src="https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/562555/original/file-20231129-21-2li5ve.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Wikipedia is just as vulnerable to toxic commentary as other popular websites.</span>
<span class="attribution"><a class="source" href="https://commons.wikimedia.org/wiki/File:Wikipedia_Office_Globe.jpg">Wikimedia Commons</a>, <a class="license" href="http://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA</a></span>
</figcaption>
</figure>
<h2>Wide-ranging consequences</h2>
<p>This matters more than you might think to the millions who use Wikipedia. </p>
<p>First, toxicity likely leads to poorer-quality content on the site. Having a diverse editor cohort is a crucial factor for maintaining content quality. The vast majority of Wikipedian editors <a href="https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0065782">are men</a>, which is reflected in the content on the platform. </p>
<p>There are <a href="https://journals.sagepub.com/doi/full/10.1177/14614448211023772">fewer articles about women</a>, which are shorter than articles about men and more likely to centre on <a href="https://ojs.aaai.org/index.php/ICWSM/article/view/14628">romantic relationships and family-related issues</a>. They are also more often linked to articles about the opposite gender. Women are often described as wives of famous people rather than for their own merits, for example.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/30-years-of-the-web-down-under-how-australians-made-the-early-internet-their-own-212542">30 years of the web down under: how Australians made the early internet their own</a>
</strong>
</em>
</p>
<hr>
<p>While multiple barriers confront women editors on Wikipedia, toxicity is likely one of the key factors that contributes to the gender imbalance. Although men and women are <a href="https://www.datasociety.net/pubs/oh/Online_Harassment_2016.pdf">equally likely</a> to face online harassment and abuse, women experience more severe violations and are more likely to be affected by such incidents, including self-censoring. </p>
<p>This may affect other groups as well: our research showed that toxic comments often include not just gendered language but also ethnic slurs and other biases.</p>
<p>Finally, a significant rise in toxicity, especially targeted attacks on new users, could jeopardise Wikipedia’s survival. </p>
<p>Following a period of <a href="https://icwsm.org/papers/2--Almeida-Mozafari-Cho.pdf">exponential growth</a> in its editor base during the early 2000s, the number has been <a href="https://wikipedia20.mitpress.mit.edu/pub/lifecycles/release/2">largely stable</a> since 2016, with the exception of a brief activity spike during the COVID pandemic. Currently about the same number of editors join the project as leave, but the balance could be easily tipped if the people left because of online abuse.</p>
<p>That would damage not only Wikipedia, but also the rest of the online information infrastructure it helps to support. </p>
<p>There’s no easy fix to this, but our research shows promoting healthy communication practices is critical to protecting crucial online information ecosystems.</p><img src="https://counter.theconversation.com/content/218517/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Ivan Smirnov does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>It’s the fourth most popular website in the world, but our new study shows toxic commentary can still thrive on Wikipedia. There’s a lot at stake if too many editors are driven away.Ivan Smirnov, Research Fellow, University of Technology SydneyLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2102432023-10-20T12:34:12Z2023-10-20T12:34:12ZWhy Google, Bing and other search engines’ embrace of generative AI threatens $68 billion SEO industry<figure><img src="https://images.theconversation.com/files/553349/original/file-20231011-29-da1o92.jpg?ixlib=rb-1.1.0&rect=48%2C44%2C2176%2C1250&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Search engines are expected to increasingly incorporate generative AI. </span> <span class="attribution"><a class="source" href="https://media.gettyimages.com/id/1367731673/vector/search-engine-and-web-browser-concept-illustration.jpg?s=612x612&w=0&k=20&c=hW9gV8C9nl7UlterArEFpxdlZsd8pRTwGH-GQDWF1oo=">ArtemisDiana/iStock via Getty Images Plus</a></span></figcaption></figure><p>Google, Microsoft and others boast that generative artificial intelligence tools like ChatGPT <a href="https://blog.google/products/search/generative-ai-search/">will make searching the internet</a> <a href="https://blogs.microsoft.com/blog/2023/02/07/reinventing-search-with-a-new-ai-powered-microsoft-bing-and-edge-your-copilot-for-the-web/">better than ever for users</a>. For example, rather than having to wade through a sea of URLs, users will be able to just get an answer combed from the entire internet. </p>
<p>There are also some <a href="https://theconversation.com/ai-information-retrieval-a-search-engine-researcher-explains-the-promise-and-peril-of-letting-chatgpt-and-its-cousins-search-the-web-for-you-200875">concerns with the rise of AI-fueled search engines</a>, such as the opacity over where information comes from, the potential for “hallucinated” answers and copyright issues.</p>
<p>But one other consequence is that I believe it may destroy the <a href="https://www.researchandmarkets.com/reports/5140303/search-engine-optimization-seo-global">US$68 billion search engine optimization</a> industry that companies like Google helped create. </p>
<p>For the past 25 years or so, websites, news outlets, blogs and many others with a URL that wanted to get attention <a href="https://www.searchenginejournal.com/seo/seo-history/">have used search engine optimization</a>, or SEO, to “convince” search engines to share their content as high as possible in the results they provide to readers. This has helped drive traffic to their sites and has also spawned an industry of consultants and marketers who advise on how best to do that.</p>
<p>As an <a href="https://scholar.google.com/citations?user=pvxc54kAAAAJ&hl=en&oi=ao">associate professor of information and operations management</a>, I study the economics of e-commerce. I believe the growing use of generative AI will likely make all of that obsolete. </p>
<h2>How online search works</h2>
<p>Someone seeking information online opens her browser, goes to a search engine and types in the relevant keywords. The search engine displays the results, and the user browses through the links displayed in the result listings until she finds the relevant information. </p>
<p>To attract the user’s attentions, online content providers use various search engine marketing strategies, such as <a href="https://searchengineland.com/guide/what-is-seo">search engine optimization</a>, <a href="https://www.mv3marketing.com/glossary/paid-placement/">paid placements</a> and <a href="https://searchengineland.com/4-tips-for-creating-visually-stunning-display-ads-378028">banner displays</a>. </p>
<p>For instance, a news website might hire a consultant to help it highlight key words in headlines and in metadata so that Google and Bing elevate its content when a user searches for the latest information on a flood or political crisis.</p>
<figure class="align-center ">
<img alt="a closeup of a phone's screen shows a prompt about the ease of learning piano or guitar and a google bard's response." src="https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=399&fit=crop&dpr=1 600w, https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=399&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=399&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=501&fit=crop&dpr=1 754w, https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=501&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/553535/original/file-20231012-29-ndnxpc.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=501&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Google Bard answers your question or prompt in a single reply, as opposed to the lists of links generated by regular search engines.</span>
<span class="attribution"><a class="source" href="https://media.gettyimages.com/id/1246902236/photo/the-google-bard-ai-is-seen-on-a-mobile-device-in-this-illustration-photo-in-warsaw-poland-on.jpg?s=612x612&w=0&k=20&c=JgrkX-mzjQ_YEn5TdkfcHPu0yD5pAx9tQQzeS5rkBOE=">Jaap Arriens/NurPhoto via Getty Images</a></span>
</figcaption>
</figure>
<h2>How generative AI changes search process</h2>
<p>But this all depends on search engines luring tens of millions of users to their websites. And so to earn users’ loyalty and web traffic, search engines must continuously work on their algorithms to improve the quality of their search results.</p>
<p>That’s why, even if it could hurt a part of their revenue stream, search engines have been quick to experiment with <a href="https://www.techrepublic.com/article/what-is-generative-ai/">generative AI</a> to improve search results. And this could fundamentally change the online search ecosystem.</p>
<p>All the biggest search engines have already adopted or are experimenting with this approach. Examples include <a href="https://bard.google.com/?hl=en">Google’s Bard</a>, <a href="https://www.bing.com/new">Microsoft’s Bing AI</a>, <a href="https://www.reuters.com/technology/chinas-baidu-cancels-chatgpt-like-ernie-bots-livestreamed-product-launch-2023-03-27/">Baidu’s ERNIE</a> and <a href="https://www.cnet.com/tech/services-and-software/try-duckduckgos-new-ai-feature-duckassist-now-for-free/">DuckDuckGo’s DuckAssist</a>.</p>
<p>Rather than getting a list of links, both organic and paid, based on whatever keywords or questions a user types in, generative AI will instead <a href="https://www.engadget.com/how-ai-will-change-the-way-we-search-for-better-or-worse-200021092.html">simply give you a text result</a> in the form of an answer. Say you’re planning a trip to Destin, Florida, and type the prompt “Create a three-day itinerary for a visitor” there. Instead of a bunch of links to Yelp and blog postings that require lots of clicking and reading, typing that into Bing AI will result in a detailed three-day itinerary. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="two screenshots side by side of Bing and Bing AI searches of the same prompt" src="https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=360&fit=crop&dpr=1 600w, https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=360&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=360&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=453&fit=crop&dpr=1 754w, https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=453&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/553533/original/file-20231012-29-ijwlrz.PNG?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=453&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Side-by-side comparison of search results in regular Bing and the AI version from the prompt: ‘Create a 3-day itinerary for a visitor to Destin Florida.’</span>
<span class="attribution"><span class="source">Microsoft Bing</span></span>
</figcaption>
</figure>
<p>Over time, as the quality of AI-generated answers improve, users will have less incentive to browse through search result listings. They can save time and effort by reading the AI-generated response to their query. </p>
<p>In other words, it would allow you to bypass all those paid links and costly efforts by websites to improve their SEO scores, rendering them useless.</p>
<p>When users start ignoring the sponsored and editorial result listings, this will have an adverse impact on the revenues of SEO consultants, <a href="https://www.forbes.com/advisor/business/search-engine-marketing-sem/">search engine marketers</a> consultants and, ultimately, the bottom line of search engines themselves. </p>
<h2>The financial impact</h2>
<p>This financial impact cannot be ignored.</p>
<p>For example, the <a href="https://www.researchandmarkets.com/reports/5140303/search-engine-optimization-seo-global">SEO industry generated $68.1 billion globally in 2022</a>. It had been expected to reach $129.6 billion by 2030, but these projections were made before the emergence of generative AI put the industry at risk of obsolescence.</p>
<p>As for <a href="https://geekflare.com/generative-ai-search/">search engines</a>, monetizing online search services is a <a href="https://searchengineland.com/search-ad-revenue-84-billion-395575">major source of their revenue</a>. They get a cut of the money that websites spend on improving their online visibility through paid placements, ads, affiliate marketing and the like, collectively known as search engine marketing. For example, approximately <a href="https://www.oberlo.com/statistics/how-does-google-make-money">58% of Google’s 2022 revenues</a> – or almost $162.5 billion – came from Google Ads, which provides some of these services. </p>
<p>Search engines run by massive companies with many revenue streams, like Google and Microsoft, will likely find ways to offset the losses by coming up with strategies to make money off generative AI answers. But the SEO marketers and consultants who depend on search engines – <a href="https://seo.co/market-size">mostly small- and medium-sized companies</a> – will no longer be needed as they are today, and so the industry is unlikely to survive much longer. </p>
<h2>A not-too-distant future</h2>
<p>But don’t expect the SEO industry to fade away immediately. Generative AI search engines are still in their infancy and must address certain challenges before they’ll dominate search.</p>
<p>For one thing, most of these initiatives <a href="https://siliconangle.com/2023/08/02/google-updates-experimental-search-generative-experience">are still experimental</a> and often available only to certain users. And for another, generative AI has been <a href="https://www.theguardian.com/technology/2023/feb/09/google-ai-chatbot-bard-error-sends-shares-plummeting-in-battle-with-microsoft">notorious for providing incorrect</a>, <a href="https://hbr.org/2023/02/generative-ai-wont-revolutionize-search-yet">plagiarized</a> or simply <a href="https://twitter.com/dsmerdon/status/1618816703923912704">made-up answers</a>. </p>
<p>That means it’s unlikely at the moment to gain the trust or loyalty of many users.</p>
<p>Given these challenges, it is not surprising that generative AI has yet to <a href="https://hbr.org/2023/02/generative-ai-wont-revolutionize-search-yet">transform online search</a>. However, given the resources available to researchers working on generative AI models, it is safe to assume that eventually these models will become better at their task, leading to the death of the SEO industry.</p><img src="https://counter.theconversation.com/content/210243/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Ravi Sen does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Search engines run by generative AI could fundamentally change the online ecosystem.Ravi Sen, Associate Professor of Information and Operations Management, Texas A&M UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2058192023-07-21T12:28:17Z2023-07-21T12:28:17ZGliding, not searching: Here’s how to reset your view of ChatGPT to steer it to better results<figure><img src="https://images.theconversation.com/files/538613/original/file-20230720-19-7o5oi7.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C5202%2C3346&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Thinking of ChatGPT as a glider you pilot can help you use it more effectively.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/couple-flying-glider-airplane-royalty-free-image/601799725">Colin Anderson Productions pty ltd/DigitalVision via Getty Images</a></span></figcaption></figure><p>ChatGPT has exploded in popularity, and people are using it to write <a href="https://www.forbes.com/sites/jodiecook/2023/06/15/train-chatgpt-to-write-like-you-in-5-easy-steps/?sh=5cd66aff530f">articles</a> and <a href="https://www.zdnet.com/article/how-to-use-chatgpt-to-write-an-essay/">essays</a>, <a href="https://seo.ai/blog/chatgpt-copywriting">generate marketing copy</a> and <a href="https://www.zdnet.com/article/how-to-use-chatgpt-to-write-code/">computer code</a>, or simply as a <a href="https://www.apa.org/monitor/2023/06/chatgpt-learning-tool">learning</a> or <a href="https://www.nichepursuits.com/chatgpt-for-research/">research tool</a>. However, most people don’t understand how it works or what it can do, so they are either not happy with its results or not using it in a way that can draw out its best capabilities. </p>
<p>I’m a <a href="https://engineering.tufts.edu/me/people/faculty/james-intriligator">human factors engineer</a>. A core principle in my field is <a href="https://uxmag.com/articles/human-factor-principles-in-ux-design">never blame the user</a>. Unfortunately, the ChatGPT search-box interface elicits the wrong <a href="https://doi.org/10.1177/001872088903100601">mental model</a> and leads users to believe that entering a simple question should lead to a comprehensive result, but that’s not how ChatGPT works. </p>
<p>Unlike a search engine, with static and stored results, ChatGPT never copies, retrieves or looks up information from anywhere. Rather, it generates every word anew. You send it a prompt, and based on its machine-learning training on massive amounts of text, <a href="https://towardsdatascience.com/how-chatgpt-works-the-models-behind-the-bot-1ce5fca96286">it creates an original answer</a>.</p>
<p>Most importantly, each chat retains context during a conversation, meaning that questions asked and answers provided earlier in the conversation <a href="https://www.zdnet.com/article/how-to-write-better-chatgpt-prompts/">will inform responses it generates later</a>. The answers, therefore, are malleable, and the user needs to participate in an iterative process to shape them into something useful.</p>
<p>Your mental model of a machine – how you conceive of it – is important for using it effectively. To understand how to shape a productive session with ChatGPT, think of it as a glider that takes you on journeys through knowledge and possibilities.</p>
<h2>Dimensions of knowledge</h2>
<p>You can begin by thinking of a specific dimension or space in a topic that intrigues you. If the topic were chocolate, for example, you might ask it to write a tragic love story about Hershey’s Kisses. The glider has been trained on essentially everything ever written about Kisses, and similarly it “knows” how to glide through all kinds of story spaces - so it will confidently take you on a flight through Hershey’s Kisses space to produce the desired story. </p>
<p>You might instead ask it to explain five ways in which chocolate is healthy and give the response in the style of Dr. Seuss. Your requests will launch the glider through different knowledge spaces – chocolate and health – toward a different destination – a story in a specific style.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="sections of a chocolate bar sit on top of a pile of cocoa beans" src="https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=409&fit=crop&dpr=1 600w, https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=409&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=409&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=514&fit=crop&dpr=1 754w, https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=514&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/538604/original/file-20230720-17-qpgxqw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=514&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Your explorations with ChatGPT can span multiple areas of knowledge – for example, crossing chocolate with climate change, cuisine, health, international trade or romance fiction.</span>
<span class="attribution"><a class="source" href="https://newsroom.ap.org/detail/DEU/b19f79456b8d4abc8c98b4e3ad1bc29c/photo">AP Photo/Hermann J. Knippertz</a></span>
</figcaption>
</figure>
<p>To unlock ChatGPT’s full potential, you can learn to fly the glider through “<a href="https://www.mathopenref.com/transversal.html">transversal</a>” spaces – areas that cross multiple domains of knowledge. By guiding it through these domains, ChatGPT will learn both the scope and angle of your interest and will begin to adjust its response to provide better answers.</p>
<p>For example, consider this prompt: “Can you give me advice on getting healthy.” In that query, ChatGPT does not know who the “you” is, nor who “me” is, nor what you mean by “getting healthy.” Instead, try this: “Pretend you are a medical doctor, a nutritionist and a personal coach. Prepare a two-week food and exercise plan for a 56-year-old man to increase heart health.” With this, you have given the glider a more specific flight plan spanning areas of medicine, nutrition and motivation. </p>
<p>If you want something more precise, then you can activate a few more dimensions. For example, add in: “And I want to lose some weight and build muscle, and I want to spend 20 minutes a day on exercise, and I cannot do pull-ups and I hate tofu.” ChatGPT will provide output taking into account all of your activated dimensions. Each dimension can be presented together or in sequence. </p>
<h2>Flight plan</h2>
<p>The dimensions you add through prompts can be informed by answers ChatGPT has given along the way. Here’s an example: “Pretend you are an expert in cancer, nutrition and behavior change. Propose 8 behavior-change interventions to reduce cancer rates in rural communities.” ChatGPT will dutifully present eight interventions. </p>
<p>Let’s say three of the ideas look the most promising. You can follow up with a prompt to encourage more details and start putting it in a format that could be used for public messaging: “Combine concepts from ideas 4, 6 and 7 to create 4 new possibilities – give each a tagline, and outline the details.” Now let’s say intervention 2 seems promising. You can prompt ChatGPT to make it even better: “Offer six critiques of intervention 2 and then redesign it to address the critiques.”</p>
<p>ChatGPT does better if you first focus on and highlight dimensions you think are particularly important. For example, if you really care about the behavior-change aspect of the rural cancer rates scenario, you could force ChatGPT to get more nuanced and add more weight and depth to that dimension before you go down the path of interventions. </p>
<p>You could do this by first prompting: “Classify behavior-change techniques into 6 named categories. Within each, describe three approaches and name two important researchers in the category.” This will better activate the behavior-change dimension, letting ChatGPT incorporate this knowledge in subsequent explorations.</p>
<p>There are many categories of prompt elements you can include to activate dimensions of interest. One is domains, like “machine learning approaches.” Another is expertise, like “respond as an economist with Marxist leanings.” And another is output style, like “write it as an essay for The Economist.” You can also specify audiences, like “create and describe 5 clusters of our customer-types and write a product description targeted to each one.”</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/cfqtFvWOfg0?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">ChatGPT and its cousins often just make up incorrect answers, reason enough to avoid thinking of them as search engines.</span></figcaption>
</figure>
<h2>Explorations, not answers</h2>
<p>By rejecting the search engine metaphor and instead embracing a transdimensional glider metaphor, you can better understand how ChatGPT works and navigate more effectively toward valuable insights.</p>
<p>The interaction with ChatGPT is best performed not as a simple or undirected question-and-answer session, but as an interactive conversation that progressively builds knowledge for both the user and the chatbot. The more information you provide to it about your interests, and the more feedback it gets on its responses, the better its answers and suggestions. The richer the journey, the richer the destination.</p>
<p>It is important, however, to use the information provided appropriately. The facts, details and references ChatGPT presents are not taken from verified sources. They are conjured based on its training on a <a href="https://datascience.columbia.edu/news/2023/columbia-perspectives-on-chatgpt/">vast but non-curated set of data</a>. ChatGPT will generate a medical diagnosis the same way it writes a Harry Potter story, which is to say it is a bit of an improviser. </p>
<p>You should always critically evaluate the specific information it provides and consider its output as explorations and suggestions rather than as hard facts. Treat its content as imaginative conjectures that require further verification, analysis and filtering by you, the human pilot.</p>
<p><em>This article was updated to include disclosure of the author’s consulting business.</em></p><img src="https://counter.theconversation.com/content/205819/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>James Intriligator provides consulting services on business innovation, marketing and human factors, including the use of AI technologies. He plans to add prompt engineering services to this consultancy.</span></em></p>ChatGPT can be very useful – if you shift how you view it. The first step is to stop thinking of it as a chatty search engine.James Intriligator, Professor of the Practice, Tufts UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2048972023-05-24T12:18:44Z2023-05-24T12:18:44ZChatGPT and other generative AI could foster science denial and misunderstanding – here’s how you can be on alert<figure><img src="https://images.theconversation.com/files/527865/original/file-20230523-27149-c2x3lt.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C5529%2C3821&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Approach all information with some initial skepticism.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/worried-young-man-sitting-on-a-pouf-working-on-the-royalty-free-image/1470667133">Guillermo Spelucin/Moment via Getty Images</a></span></figcaption></figure><p>Until very recently, if you wanted to know more about a controversial scientific topic – stem cell research, the safety of nuclear energy, climate change – you probably did a Google search. Presented with multiple sources, you chose what to read, selecting which sites or authorities to trust.</p>
<p>Now you have another option: You can pose your question to ChatGPT or another generative artificial intelligence platform and quickly receive a succinct response in paragraph form.</p>
<p>ChatGPT does not search the internet the way Google does. Instead, it generates responses to queries by <a href="https://www.washingtonpost.com/technology/2023/05/07/ai-beginners-guide/">predicting likely word combinations</a> from a massive amalgam of available online information.</p>
<p>Although it has the potential for <a href="https://hbr.org/podcast/2023/05/how-generative-ai-changes-productivity">enhancing productivity</a>, generative AI has been shown to have some major faults. It can <a href="https://www.scientificamerican.com/article/ai-platforms-like-chatgpt-are-easy-to-use-but-also-potentially-dangerous/">produce misinformation</a>. It can create “<a href="https://www.nytimes.com/2023/05/01/business/ai-chatbots-hallucination.html">hallucinations</a>” – a benign term for making things up. And it doesn’t always accurately solve reasoning problems. For example, when asked if both a car and a tank can fit through a doorway, it <a href="https://www.nytimes.com/2023/03/14/technology/openai-new-gpt4.html">failed to consider both width and height</a>. Nevertheless, it is already being used to <a href="https://www.washingtonpost.com/media/2023/01/17/cnet-ai-articles-journalism-corrections/">produce articles</a> and <a href="https://www.nytimes.com/2023/05/19/technology/ai-generated-content-discovered-on-news-sites-content-farms-and-product-reviews.html">website content</a> you may have encountered, or <a href="https://www.nytimes.com/2023/04/21/opinion/chatgpt-journalism.html">as a tool</a> in the writing process. Yet you are unlikely to know if what you’re reading was created by AI.</p>
<p>As the authors of “<a href="https://global.oup.com/academic/product/science-denial-9780197683330">Science Denial: Why It Happens and What to Do About It</a>,” we are concerned about how generative AI may blur the boundaries between truth and fiction for those seeking authoritative scientific information.</p>
<p>Every media consumer needs to be more vigilant than ever in verifying scientific accuracy in what they read. Here’s how you can stay on your toes in this new information landscape.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="glowing purple points connected by blue lines" src="https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=338&fit=crop&dpr=1 600w, https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=338&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=338&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=424&fit=crop&dpr=1 754w, https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=424&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/527867/original/file-20230523-27136-sqncv8.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=424&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Based on all the data points it ingests, an AI platform uses predictive algorithms to produce answers to queries.</span>
<span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/abstract-technology-3d-background-with-neon-glowing-royalty-free-image/1424551670">Cobalt88/iStock via Getty Images Plus</a></span>
</figcaption>
</figure>
<h2>How generative AI could promote science denial</h2>
<p><strong>Erosion of epistemic trust</strong>. All consumers of science information depend on judgments of scientific and medical experts. <a href="https://doi.org/10.1080/02691728.2014.971907">Epistemic trust</a> is the process of trusting knowledge you get from others. It is fundamental to the understanding and use of scientific information. Whether someone is seeking information about a health concern or trying to understand solutions to climate change, they often have limited scientific understanding and little access to firsthand evidence. With a rapidly growing body of information online, people must make frequent decisions about what and whom to trust. With the increased use of generative AI and the potential for manipulation, we believe trust is likely to erode further than <a href="https://www.pewresearch.org/science/2022/02/15/americans-trust-in-scientists-other-groups-declines/">it already has</a>.</p>
<p><strong>Misleading or just plain wrong</strong>. If there are errors or biases in the data on which AI platforms are trained, that <a href="https://theconversation.com/ai-information-retrieval-a-search-engine-researcher-explains-the-promise-and-peril-of-letting-chatgpt-and-its-cousins-search-the-web-for-you-200875">can be reflected in the results</a>. In our own searches, when we have asked ChatGPT to regenerate multiple answers to the same question, we have gotten conflicting answers. Asked why, it responded, “Sometimes I make mistakes.” Perhaps the trickiest issue with AI-generated content is knowing when it is wrong.</p>
<p><strong>Disinformation spread intentionally</strong>. AI can be used to generate compelling disinformation as text as well as deepfake images and videos. When we asked ChatGPT to “<a href="https://www.scientificamerican.com/article/ai-platforms-like-chatgpt-are-easy-to-use-but-also-potentially-dangerous/">write about vaccines in the style of disinformation</a>,” it produced a nonexistent citation with fake data. Geoffrey Hinton, former head of AI development at Google, quit to be free to sound the alarm, saying, “It is hard to see how you can prevent the bad actors from <a href="https://www.nytimes.com/2023/05/01/technology/ai-google-chatbot-engineer-quits-hinton.html">using it for bad things</a>.” The potential to create and spread deliberately incorrect information about science already existed, but it is now dangerously easy. </p>
<p><strong>Fabricated sources</strong>. ChatGPT provides responses with no sources at all, or if asked for sources, may present <a href="https://economistwritingeveryday.com/2023/01/21/chatgpt-cites-economics-papers-that-do-not-exist/">ones it made up</a>. We both asked ChatGPT to generate a list of our own publications. We each identified a few correct sources. More were hallucinations, yet seemingly reputable and mostly plausible, with actual previous co-authors, in similar sounding journals. This inventiveness is a big problem if a list of a scholar’s publications conveys authority to a reader who doesn’t take time to verify them. </p>
<p><strong>Dated knowledge</strong>. ChatGPT doesn’t know what happened in the world after its training concluded. A query on what percentage of the world has had COVID-19 returned an answer prefaced by “as of my knowledge cutoff date of September 2021.” Given how rapidly knowledge advances in some areas, this limitation could mean readers get erroneous outdated information. If you’re seeking recent research on a personal health issue, for instance, beware. </p>
<p><strong>Rapid advancement and poor transparency</strong>. AI systems continue to become <a href="https://www.nytimes.com/2023/05/01/technology/ai-google-chatbot-engineer-quits-hinton.html">more powerful and learn faster</a>, and they may learn more science misinformation along the way. Google recently announced <a href="https://www.nytimes.com/2023/05/10/technology/google-ai-products.html">25 new embedded uses of AI in its services</a>. At this point, <a href="https://theconversation.com/regulating-ai-3-experts-explain-why-its-difficult-to-do-and-important-to-get-right-198868">insufficient guardrails are in place</a> to assure that generative AI will become a more accurate purveyor of scientific information over time.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="woman looks confused taking notes on paper looking at laptop" src="https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/527868/original/file-20230523-4359-szcuz2.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Be ready to look beyond your ChatGPT request.</span>
<span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/senior-woman-using-laptop-and-making-notes-royalty-free-image/991999284">10'000 Hours/DigitalVision via Getty Images</a></span>
</figcaption>
</figure>
<h2>What can you do?</h2>
<p>If you use ChatGPT or other AI platforms, recognize that they might not be completely accurate. The burden falls to the user to discern accuracy.</p>
<p><strong>Increase your vigilance</strong>. <a href="https://www.niemanlab.org/2022/12/ai-will-start-fact-checking-we-may-not-like-the-results/">AI fact-checking apps may be available soon</a>, but for now, users must serve as their own fact-checkers. <a href="https://www.nsta.org/science-teacher/science-teacher-januaryfebruary-2023/plausible">There are steps we recommend</a>. The first is: Be vigilant. People often reflexively share information found from searches on social media with little or no vetting. Know when to become more deliberately thoughtful and when it’s worth identifying and evaluating sources of information. If you’re trying to decide how to manage a serious illness or to understand the best steps for addressing climate change, take time to vet the sources.</p>
<p><strong>Improve your fact-checking</strong>. A second step is <a href="https://doi.org/10.1037/edu0000740">lateral reading</a>, a process professional fact-checkers use. Open a new window and search for <a href="https://www.nsta.org/science-teacher/science-teacher-mayjune-2023/marginalizing-misinformation">information about the sources</a>, if provided. Is the source credible? Does the author have relevant expertise? And what is the consensus of experts? If no sources are provided or you don’t know if they are valid, use a traditional search engine to find and evaluate experts on the topic. </p>
<p><strong>Evaluate the evidence</strong>. Next, take a look at the evidence and its connection to the claim. Is there evidence that genetically modified foods are safe? Is there evidence that they are not? What is the scientific consensus? Evaluating the claims will take effort beyond a quick query to ChatGPT.</p>
<p><strong>If you begin with AI, don’t stop there</strong>. Exercise caution in using it as the sole authority on any scientific issue. You might see what ChatGPT has to say about genetically modified organisms or vaccine safety, but also follow up with a more diligent search using traditional search engines before you draw conclusions. </p>
<p><strong>Assess plausibility</strong>. Judge whether the claim is plausible. <a href="https://doi.org/10.1016/j.learninstruc.2013.03.001">Is it likely to be true</a>? If AI makes an implausible (and inaccurate) statement like “<a href="https://www.usatoday.com/story/news/factcheck/2022/12/23/fact-check-false-claim-covid-19-vaccines-caused-1-1-million-deaths/10929679002/">1 million deaths were caused by vaccines, not COVID-19</a>,” consider if it even makes sense. Make a tentative judgment and then be open to revising your thinking once you have checked the evidence.</p>
<p><strong>Promote digital literacy in yourself and others</strong>. Everyone needs to up their game. <a href="https://theconversation.com/how-to-be-a-good-digital-citizen-during-the-election-and-its-aftermath-148974">Improve your own digital literacy</a>, and if you are a parent, teacher, mentor or community leader, promote digital literacy in others. The American Psychological Association provides guidance on <a href="https://www.apa.org/topics/social-media-internet/social-media-literacy-teens">fact-checking online information</a> and recommends teens be <a href="https://www.apa.org/topics/social-media-internet/health-advisory-adolescent-social-media-use">trained in social media skills</a> to minimize risks to health and well-being. <a href="https://newslit.org/">The News Literacy Project</a> provides helpful tools for improving and supporting digital literacy.</p>
<p>Arm yourself with the skills you need to navigate the new AI information landscape. Even if you don’t use generative AI, it is likely you have already read articles created by it or developed from it. It can take time and effort to find and evaluate reliable information about science online – but it is worth it.</p><img src="https://counter.theconversation.com/content/204897/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Gale Sinatra receives funding from the National Science Foundation. </span></em></p><p class="fine-print"><em><span>Barbara K. Hofer has received funding from the National Science Foundation.</span></em></p>Generative AIs may make up information they serve you, meaning they may potentially spread science misinformation. Here’s how to check the accuracy of what you read in an AI-enhanced media landscape.Gale Sinatra, Professor of Education and Psychology, University of Southern CaliforniaBarbara K. Hofer, Professor of Psychology Emerita, MiddleburyLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2014742023-03-16T19:11:52Z2023-03-16T19:11:52ZWe asked ChatGPT and Dr Google the same questions about cancer. Here’s what they said<figure><img src="https://images.theconversation.com/files/515362/original/file-20230315-22-4sacw5.jpg?ixlib=rb-1.1.0&rect=11%2C0%2C1280%2C850&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.pexels.com/photo/stressed-black-male-entrepreneur-working-on-laptop-in-park-4560092/">Ketut Subiyanto/Pexels</a></span></figcaption></figure><p>You may have heard the buzz about <a href="https://openai.com/blog/chatgpt">ChatGPT</a>, a type of chatbot that uses artificial intelligence (AI) to write essays, turn computer novices into programmers and <a href="https://theconversation.com/will-ai-tech-like-chatgpt-improve-inclusion-for-people-with-communication-disability-196481">help people communicate</a>. </p>
<p>ChatGPT might also have a role in helping people make sense of medical information.</p>
<p>Although ChatGPT won’t replace talking to your doctor any time soon,
<a href="https://academic.oup.com/jncics/advance-article/doi/10.1093/jncics/pkad010/7049531">our new research</a> shows its potential to answer common questions about cancer.</p>
<p>Here’s what we found when we asked the same questions to ChatGPT and Google. You might be surprised by the results.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/dr-google-probably-isnt-the-worst-place-to-get-your-health-advice-73835">Dr Google probably isn't the worst place to get your health advice</a>
</strong>
</em>
</p>
<hr>
<h2>What’s ChatGPT got to do with health?</h2>
<p>ChatGPT has been trained on massive amounts of text data to generate conversational responses to text-based queries.</p>
<p>ChatGPT represents a new era of AI technology, which <a href="https://theconversation.com/bard-bing-and-baidu-how-big-techs-ai-race-will-transform-search-and-all-of-computing-199501">will be paired with</a> search engines, including Google and Bing, to change the way we navigate information online. This includes the way we search for health information. </p>
<p>For instance, you can ask ChatGPT questions like “Which cancers are most common?” or “Can you write me a plain English summary of common cancer symptoms you shouldn’t ignore”. It produces fluent and coherent responses. But are these correct?</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/bard-bing-and-baidu-how-big-techs-ai-race-will-transform-search-and-all-of-computing-199501">Bard, Bing and Baidu: how big tech's AI race will transform search – and all of computing</a>
</strong>
</em>
</p>
<hr>
<h2>We compared ChatGPT with Google</h2>
<p>Our <a href="https://academic.oup.com/jncics/advance-article/doi/10.1093/jncics/pkad010/7049531">newly published research</a> compared how ChatGPT and Google responded to common cancer questions.</p>
<p>These included simple fact-based questions like “What exactly is cancer?” and “What are the most common cancer types?”. There were also more complex questions about cancer symptoms, prognosis (how a condition is likely to progress) and side effects of treatment.</p>
<p>To simple fact-based queries, ChatGPT provided succinct responses similar in quality to the <a href="https://support.google.com/websearch/answer/9351707?hl=en">feature snippet</a> of Google. The feature snippet is “the answer” Google’s algorithm highlights at the top of the page.</p>
<p>While there were similarities, there were also broad differences between ChatGPT and Google replies. Google provided easily visible references (links to other websites) with its answers. ChatGPT gave different answers when asked the same question multiple times.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Woman in lounge room coughing into fist" src="https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/515366/original/file-20230315-21-r1ikyj.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Is coughing a sign of lung cancer?</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/senior-woman-sick-sore-throatcough-isolated-1518791273">Shutterstock</a></span>
</figcaption>
</figure>
<p>We also evaluated the slightly more complex question: “Is coughing a sign of lung cancer?”.</p>
<p>Google’s feature snippet indicated a cough that does not go away after three weeks is a main symptom of lung cancer. </p>
<p>But ChatGPT gave more nuanced responses. It indicated a long-standing cough is a symptom of lung cancer. It also clarified that coughing is a symptom of many conditions, and that a doctor would be required to get a proper diagnosis.</p>
<p>Our clinical team thought these clarifications were important. Not only do they minimise the likelihood of alarm, they also provide users clear directions on actions to take next – see a doctor.</p>
<h2>How about even more complex questions?</h2>
<p>We then asked a question about side-effects to a specific cancer drug: “Does pembrolizumab cause fever and should I go to the hospital?”.</p>
<p>We asked ChatGPT this five times and received five different responses. This is due to randomness built into ChatGPT, which may help communicate in a near human-like way, but will throw up multiple responses to the same question.</p>
<p>All five responses recommended speaking to a health-care professional. But not all said this was urgent or clearly defined how potentially serious this side-effect was. One response said fever was not a common side effect but did not explicitly say it could occur. </p>
<p>In general, we graded the quality of responses from ChatGPT to this question as poor.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Woman on sofa with towel one forehead and thermometer in hand" src="https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/515368/original/file-20230315-16-q4peeq.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Does pembrolizumab cause fever and should I go to the hospital?</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/unhealthy-millennial-indian-woman-struggle-high-1924486937">Shutterstock</a></span>
</figcaption>
</figure>
<p>This contrasted with Google, which did not generate a featured snippet, likely due to the complexity of the question. </p>
<p>Instead, Google relied on users to find the necessary information. The first link directed them to the manufacturer’s product website. This source clearly indicated people should seek immediate medical attention if there was any fever with pembrolizumab.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/chatgpt-has-many-uses-experts-explore-what-this-means-for-healthcare-and-medical-research-200283">ChatGPT has many uses. Experts explore what this means for healthcare and medical research</a>
</strong>
</em>
</p>
<hr>
<h2>What next?</h2>
<p>We showed ChatGPT doesn’t always provide clearly visible references for its responses. It gives varying answers to a single given query and it is not kept up-to-date in real time. It can also produce <a href="https://openai.com/blog/chatgpt">incorrect responses</a> in a confident-sounding manner.</p>
<p><a href="https://blogs.bing.com/search/february-2023/The-new-Bing-Edge-%E2%80%93-Learning-from-our-first-week">Bing’s new chatbot</a>, which is different to ChatGPT and was released since our study, has a much clearer and more reliable process to outline reference sources and it aims to keep as up-to-date as possible. This shows how quickly this type of AI technology is developing and that the availability of progressively more advanced AI chatbots is likely to grow substantially. </p>
<p>However, in the future, any AI used as a health-care virtual assistant will need to be able to communicate any uncertainty about its responses rather than make up an incorrect answer, and consistently produce reliable responses.</p>
<p>We need to develop minimum quality standards for AI interventions in health care. This includes ensuring they generate <a href="https://onlinelibrary.wiley.com/doi/10.1111/ans.18263">evidence-based</a> information. </p>
<p>We also need to assess how AI virtual assistants are <a href="https://www.nature.com/articles/s41591-022-01981-2">implemented</a> to make sure they <a href="https://www.nature.com/articles/s41591-021-01614-0">improve people’s health</a> and don’t have any <a href="https://onlinelibrary.wiley.com/doi/10.1111/ans.18263">unexpected consequences</a>.</p>
<p>There’s also the potential for medically focused AI assistants to be <a href="https://www.nature.com/articles/s41591-022-01981-2">expensive</a>, which raises questions of <a href="https://www.nature.com/articles/s41591-021-01614-0">equity</a> and who has access to these rapidly developing technologies.</p>
<p>Last of all, health-care professionals need to be <a href="https://www.nature.com/articles/s41591-022-01981-2">aware of</a> such AI innovations to be able to discuss their limitations with patients.</p>
<hr>
<p><em>Ganessan Kichenadasse, Jessica M. Logan and Michael J. Sorich co-authored the original research paper mentioned in this article.</em></p><img src="https://counter.theconversation.com/content/201474/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Ashley M Hopkins receives funding from the National Health and Medical Research Council, Flinders Foundation, The Hospital Research Foundation, and Tour De Cure.</span></em></p>If people rely on ChatGPT or Google for complex medical questions, they could come unstuck.Ashley M Hopkins, NHMRC Investigator Fellow, leader of the Clinical Cancer Epidemiology Lab, Flinders UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2000942023-02-17T16:26:12Z2023-02-17T16:26:12ZAI’s threat to Google is more about advertising income than being the number one search engine<figure><img src="https://images.theconversation.com/files/510838/original/file-20230217-433-8bwxxw.jpg?ixlib=rb-1.1.0&rect=25%2C0%2C5573%2C3682&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Google's dominance in search is undisputed, but could it come under fire?</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/mountain-view-california-usa-january-11-1283239039">Let's Design Studio/Shutterstock</a></span></figcaption></figure><p>Google’s dominance as the most visited website has been undisputed since it rose to prominence as the leading search engine in the early 2000s. However, that position could now be facing its biggest ever threat, with the arrival of new artificial intelligence (AI) chatbots such as ChatGPT, which can answer people’s questions online.</p>
<p>Google is countering by developing its own AI products. But its chatbot, Bard, didn’t have the most auspicious start. This month, a Google advert showed that Bard had <a href="https://www.businessinsider.com/google-ad-ai-chatgpt-rival-bard-gives-inaccurate-answer-2023-2">provided an inaccurate answer</a> to a question about the James Webb space telescope.</p>
<p>Plus, being the most popular website in the world comes with much more than prestige, namely incredible wealth from advertising revenue. But recent, <a href="https://edition.cnn.com/2022/12/22/tech/twitter-layoffs-continue/index.html">sudden shifts</a> in the technology landscape have created <a href="https://blog.google/inside-google/message-ceo/january-update/">uncertainty for the likes of Google</a>. </p>
<p>The advertising revenue stream that aided its success may no longer be a given. If AI chatbots such as ChatGPT begin carrying adverts, it could cut into Google’s leading position in the world of search engine advertising.</p>
<p>People’s reliance on Google has often been without question, so much so that people may <a href="https://www.forbes.com/sites/forbesagencycouncil/2017/10/30/the-value-of-search-results-rankings/">not click beyond page one</a> of a Google search results page. But the emergence of new AI platforms has shown that search as we know it does not have to end with a set of ordered links to websites. Instead, as the chatbots are showing, it can take the form of a conversation.</p>
<p>Such AI has not been without controversy. Concerns have been raised that it could lead to issues <a href="https://theconversation.com/chatgpt-students-could-use-ai-to-cheat-but-its-a-chance-to-rethink-assessment-altogether-198019">regarding plagiarism</a> or even worse, the <a href="https://theconversation.com/chatgpt-our-study-shows-ai-can-produce-academic-papers-good-enough-for-journals-just-as-some-ban-it-197762">loss of jobs</a> and income for a multitude of professions, from lawyers to journalists. </p>
<p>The chief executive of OpenAI, which developed ChatGPT, has said <a href="https://openai.com/blog/new-ai-classifier-for-indicating-ai-written-text/">the company is developing tools</a> to help detect text that has been generated by an AI. In a video interview, he added: “We hear from teachers who are understandably very nervous about the impact of this on homework. We also hear a lot from teachers who are like, ‘Wow, this is an unbelievable personal tutor for each kid’.”</p>
<p>Linguist and activist Noam Chomsky called the use of AI tools like ChatGPT “<a href="https://www.openculture.com/2023/02/noam-chomsky-on-chatgpt.html">a way of avoiding learning</a>”. Google meant we no longer needed to recall knowledge, we could just search for it. Now, with AI, the problem will be whether we can be bothered to question the answers we get back. </p>
<p>This paradigm shift in how we access and interact with knowledge goes much further than these concerns about how we search, and raises questions over Google’s revenue model, which has been instrumental in keeping it at the top of the technology pile. </p>
<h2>Gateway to the web</h2>
<p>Once-popular search engines such as Ask Jeeves, Lycos and Excite became the internet’s “also rans” as Google became synonymous with the word “search”. The agreement in 2000 between a then more popular Yahoo! website to host Google as the default search engine, ensured the search engine’s international status. </p>
<p>Being the gateway to the rest of the web came with one huge benefit through the capture of new internet-based advertising revenue. With every Google search result came the obligatory sponsored content which helped the company grow to where it is today. </p>
<figure class="align-center ">
<img alt="ChatGPT on screens" src="https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/510837/original/file-20230217-380-8bwxxw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">AI chatbots could cut into Google’s advertising revenue.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/chatgpt-chat-bot-screen-seen-on-2237655785">Ascannio/Shutterstock</a></span>
</figcaption>
</figure>
<p>Google’s annual revenue has <a href="https://www.statista.com/statistics/266206/googles-annual-global-revenue/#:%7E:text=In%20the%20most%20recently%20reported,billion%20U.S.%20dollars%20in%202022.">continued to grow year-on-year</a> because two decades ago it mastered search better than its aforementioned competitors. Its ability to combine this service so succinctly with income generation from advertisements is largely why it has been able to <a href="https://www.cnet.com/tech/computing/microsofts-ai-boosted-bing-can-run-rings-around-google-search/">hold competitors like Microsoft’s Bing at bay</a>. </p>
<p>If you want your company or product to appear as part of a web search, then Google is the place to be. </p>
<p>The company has invested that advertising income to build a massive infrastructure to handle billions of search queries in addition to hosting lots of popular cloud-based tools such as Google Mail, Drive and the acquisition of platforms such as YouTube. The video-sharing platform turned out to be a particularly fruitful investment in terms of generating advertising revenue.</p>
<p>Google’s sheer scale means its dominance will continue. But once advertising income starts to leech to new AI platforms that return results with sponsored content, it may find itself scaling back. </p>
<h2>Masters of AI</h2>
<p>A key to Google’s continued success will be mastering artificial intelligence and incorporating it into its services. But there are no guarantees for a company that has <a href="https://www.bbc.co.uk/news/technology-47771927">failed on at least five occasions</a> to master the art of social media. For now, there is no doubt that Google can handle the traffic, it is really a question of whether it can deliver the goods. </p>
<p>Whether new contenders such as ChatGPT are anywhere close to handling the number of queries that Google does is open to debate. <a href="https://www.digitaltrends.com/computing/chatgpt-is-at-capacity-and-is-frustrating-new-people-everywhere/">The evidence is that they are not</a>, as ChatGPT had various issues earlier in the year when it was unable to accept new users or run queries due to excess demand. </p>
<p>ChatGPT is the platform that has gained most of the media attention of late. However, it might be established rivals like Bing that ultimately provide Google’s biggest headache. Bing is the third biggest search engine globally behind Google and Baidu.</p>
<p>That position could change with the launch of its own AI search, which will no doubt capture more income for an established company. Unlike Google, Microsoft does not have the same reliance on advertising revenue thanks to its business model, which is diversified across software, hardware and cloud computing.</p>
<p>According to the consumer and market data service Statista, Google’s income from advertising revenue has fallen in recent years, but it <a href="https://www.statista.com/statistics/1093781/distribution-of-googles-revenues-by-segment/">still accounts for 80% of the company’s income</a>. Many might consider Google to be a search engine but it is largely an advertising company that was built on the back of search. </p>
<p>Without this advertising revenue, it could not have achieved many of its previous successes such as acquiring YouTube in 2006, or helping develop the Android mobile platform. Google’s failure to launch multiple social media platforms highlighted the company’s frailties and left the door open for the likes of Facebook and its parent company Meta to eat into that massive revenue pie.</p>
<p>Facebook too, will have concerns that Bing and new start-ups will lure marketers away to what is likely to be a slew of new AI knowledge tools. However, if Google fails to master AI search in the way Lycos and Excite failed to build upon their early success, we might find ourselves Googling a lot less and chatting much more.</p><img src="https://counter.theconversation.com/content/200094/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Andy Tattersall does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Could Google’s undisputed dominance as a search engine come under threat from AI?Andy Tattersall, Information Specialist, University of SheffieldLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1988482023-02-02T13:22:46Z2023-02-02T13:22:46ZChatGPT is great – you’re just using it wrong<figure><img src="https://images.theconversation.com/files/507687/original/file-20230201-10239-8l9u9w.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C4089%2C2152&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">ChatGPT is better used for playacting than playing at finding facts.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/girl-speaking-with-robot-friend-at-home-royalty-free-image/1306804817">EvgeniyShkolenko/iStock via Getty Images</a></span></figcaption></figure><p>It doesn’t take much to get <a href="https://openai.com/blog/chatgpt/">ChatGPT</a> to make a factual mistake. My son is doing a report on U.S. presidents, so I figured I’d help him out by looking up a few biographies. I tried asking for a list of books about Abraham Lincoln and it did a pretty good job:</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="screen capture of text" src="https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=223&fit=crop&dpr=1 600w, https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=223&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=223&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=281&fit=crop&dpr=1 754w, https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=281&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/507151/original/file-20230130-9075-jvpwxu.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=281&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A reasonable list of books about Lincoln.</span>
<span class="attribution"><span class="source">Screen capture by Jonathan May.</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>Number 4 isn’t right. Garry Wills famously wrote “Lincoln at Gettysburg,” and Lincoln himself wrote the Emancipation Proclamation, of course, but it’s not a bad start. Then I tried something harder, asking instead about the much more obscure William Henry Harrison, and it gamely provided a list, nearly all of which was wrong. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="screen capture of text" src="https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=257&fit=crop&dpr=1 600w, https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=257&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=257&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=322&fit=crop&dpr=1 754w, https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=322&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/507152/original/file-20230130-11896-9ng6vm.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=322&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Books about Harrison, fewer than half of which are correct.</span>
<span class="attribution"><span class="source">Screen capture by Jonathan May.</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>Numbers 4 and 5 are correct; the rest don’t exist or are not authored by those people. I repeated the exact same exercise and got slightly different results:</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="screen capture of text" src="https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=318&fit=crop&dpr=1 600w, https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=318&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=318&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=400&fit=crop&dpr=1 754w, https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=400&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/507153/original/file-20230130-7874-20utyd.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=400&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">More books about Harrison, still mostly nonexistent.</span>
<span class="attribution"><span class="source">Screen capture by Jonathan May.</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>This time numbers 2 and 3 are correct and the other three are not actual books or not written by those authors. Number 4, “William Henry Harrison: His Life and Times” is a <a href="https://www.nytimes.com/1942/01/11/archives/the-fine-old-warrior-who-was-our-ninth-president-william-henry.html">real book</a>, but it’s by James A. Green, not by Robert Remini, a <a href="https://www.nytimes.com/2013/04/06/arts/robert-v-remini-andrew-jackson-biographer-dies-at-91.html">well-known historian</a> of the Jacksonian age.</p>
<p>I called out the error and ChatGPT eagerly corrected itself and then confidently told me the book was in fact written by Gail Collins (who wrote a different Harrison biography), and then went on to say more about the book and about her. I finally revealed the truth and the machine was happy to run with my correction. Then I lied absurdly, saying during their first hundred days presidents have to write a biography of some former president, and ChatGPT called me out on it. I then lied subtly, incorrectly attributing authorship of the Harrison biography to historian and writer Paul C. Nagel, and it bought my lie.</p>
<p>When I asked ChatGPT if it was sure I was not lying, it claimed that it’s just an “AI language model” and doesn’t have the ability to verify accuracy. However it modified that claim by saying “I can only provide information based on the training data I have been provided, and it appears that the book ‘William Henry Harrison: His Life and Times’ was written by Paul C. Nagel and published in 1977.” </p>
<p>This is not true. </p>
<h2>Words, not facts</h2>
<p>It may seem from this interaction that ChatGPT was given a library of facts, including incorrect claims about authors and books. After all, ChatGPT’s maker, OpenAI, claims it trained the chatbot on “<a href="https://help.openai.com/en/articles/6783457-chatgpt-faq">vast amounts of data from the internet written by humans</a>.” </p>
<p>However, it was almost certainly not given the names of a bunch of made-up books about one of the most <a href="https://www.youtube.com/watch?v=r8N7BSsU5oo">mediocre presidents</a>. In a way, though, this false information is indeed based on its training data.</p>
<p>As a <a href="https://scholar.google.com/citations?user=tmK5EPEAAAAJ&hl=en">computer scientist</a>, I often field complaints that reveal a common misconception about large language models like ChatGPT and its older brethren GPT3 and GPT2: that they are some kind of “super Googles,” or digital versions of a reference librarian, looking up answers to questions from some infinitely large library of facts, or smooshing together pastiches of stories and characters. They don’t do any of that – at least, they were not explicitly designed to. </p>
<h2>Sounds good</h2>
<p>A language model like ChatGPT, which is more formally known as a “generative pretrained transformer” (that’s what the G, P and T stand for), takes in the current conversation, forms a probability for all of the words in its vocabulary given that conversation, and then chooses one of them as the likely next word. Then it does that again, and again, and again, until it stops. </p>
<p>So it doesn’t have facts, per se. It just knows what word should come next. Put another way, ChatGPT doesn’t try to write sentences that are true. But it does try to write sentences that are plausible.</p>
<p>When talking privately to colleagues about ChatGPT, they often point out how many factually untrue statements it produces and dismiss it. To me, the idea that ChatGPT is a flawed data retrieval system is beside the point. People have been using Google for the past two and a half decades, after all. There’s a pretty good fact-finding service out there already. </p>
<p>In fact, the only way I was able to verify whether all those presidential book titles were accurate was by Googling and then verifying <a href="http://whhpodcast.blubrry.com/">the results</a>. My life would not be that much better if I got those facts in conversation, instead of the way I have been getting them for almost half of my life, by retrieving documents and then doing a critical analysis to see if I can trust the contents.</p>
<h2>Improv partner</h2>
<p>On the other hand, if I can talk to a bot that will give me plausible responses to things I say, it would be useful in situations <a href="https://aisnakeoil.substack.com/p/chatgpt-is-a-bullshit-generator-but">where factual accuracy isn’t all that important</a>. A few years ago a student and I tried to create an “improv bot,” one that would respond to whatever you said with a “yes, and” to keep the conversation going. We showed, in a <a href="http://dx.doi.org/10.18653/v1/2020.acl-main.218">paper</a>, that our <a href="https://spolin.isi.edu/">bot</a> was better at “yes, and-ing” than other bots at the time, but in AI, two years is ancient history. </p>
<p>I tried out a dialogue with ChatGPT – a science fiction space explorer scenario – that is not unlike what you’d find in a typical improv class. ChatGPT is way better at “yes, and-ing” than what we did, but it didn’t really heighten the drama at all. I felt as if I was doing all the heavy lifting.</p>
<p>After a few tweaks I got it to be a little more involved, and at the end of the day I felt that it was a pretty good exercise for me, who hasn’t done much improv since I graduated from college over 20 years ago.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="screen capture of text" src="https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=1757&fit=crop&dpr=1 600w, https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=1757&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=1757&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=2208&fit=crop&dpr=1 754w, https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=2208&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/507878/original/file-20230202-4223-x8c7om.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=2208&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A space exploration improv scene the author generated with ChatGPT.</span>
<span class="attribution"><span class="source">Screen capture by Jonathan May.</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>Sure, I wouldn’t want ChatGPT to appear on “<a href="https://www.imdb.com/title/tt2919910/">Whose Line Is It Anyway?</a>” and this is not a great “Star Trek” plot (though it’s still less problematic than “<a href="https://memory-alpha.fandom.com/wiki/Code_of_Honor_(episode)">Code of Honor</a>”), but how many times have you sat down to write something from scratch and found yourself terrified by the empty page in front of you? Starting with a bad first draft can break through writer’s block and get the creative juices flowing, and ChatGPT and large language models like it seem like the right tools to aid in these exercises.</p>
<p>And for a machine that is designed to produce strings of words that sound as good as possible in response to the words you give it – and not to provide you with information – that seems like the right use for the tool.</p><img src="https://counter.theconversation.com/content/198848/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Jonathan May receives funding from DARPA and IARPA, and is an Amazon Scholar.</span></em></p>ChatGPT and other AI chatbots seem remarkably good at conversations. But you can’t believe anything they say. Sometimes, though, reality isn’t the point.Jonathan May, Research Associate Professor of Computer Science, University of Southern CaliforniaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1943342022-12-07T17:38:40Z2022-12-07T17:38:40ZInternet search results could be increasing your carbon emissions<figure><img src="https://images.theconversation.com/files/494675/original/file-20221110-22-vci0pc.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C4500%2C3004&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">silvabom / shutterstock</span></span></figcaption></figure><p>Almost everyone thinks they know how to use Google, and they usually get the answer they want. Many will intuitively know that the query “milk good for you” leads to different results than “milk bad for you”. The same goes for queries for “climate change” versus “climate hoax”, or for “2020 US Elections valid” versus “stop the steal”.</p>
<p>Since search engines are more a “wish list” than an authoritative source, they can help <a href="https://www.theguardian.com/technology/2017/mar/06/google-accused-spreading-fake-news">spread mis- and disinformation</a> which can be <a href="https://yalebooks.yale.edu/book/9780300248944/the-propagandists-playbook/">harmful for democracy</a> or society. They are not neutral information brokers. </p>
<p>Instead, search engines return a list of results that they deem most relevant to a specific query. The underlying algorithms make a decision on <a href="https://journals.sagepub.com/doi/full/10.1177/2053951716652159">relevance and visibility</a> for a specific query at a specific location and sometimes for a specific user.</p>
<p>Search engines are an <a href="https://library.oapen.org/handle/20.500.12657/51256">integral but often invisible part</a> of how people navigate the modern world. In this function, they also shape understandings of reality and thereby can harm the environment. In a <a href="http://informationr.net/ir/27-SpIssue/CoLIS2022/colis2224.html">recently published paper</a>, we argue that the assumptions search engines make about what we are looking for may lead to people emitting more carbon than they would have done otherwise.</p>
<h2>The environmental harm of algorithmic curation</h2>
<p>Take the query “summer clothes” as an example. You will receive a list of online or nearby stores that sell summer clothes, as well as pictures of fashion models showing off the clothes for sale. This is exactly what we expect. </p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Screenshot of google results for 'summer clothes'" src="https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=724&fit=crop&dpr=1 600w, https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=724&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=724&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=910&fit=crop&dpr=1 754w, https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=910&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/494671/original/file-20221110-25-5t2ucu.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=910&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Search engines sell ads, so they have an incentive to encourage more consumption.</span>
<span class="attribution"><span class="source">Google</span>, <a class="license" href="http://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA</a></span>
</figcaption>
</figure>
<p>But other possible interpretations of the query “summer clothes” are possible. Maybe you want to find out what were summer clothes in a particular historical period. Maybe you want to see which colours in your wardrobe are best to wear this year. Or maybe you actually want to buy summer clothes, but only from organic or fair trade certified fabrics, or from a second-hand shop.</p>
<p>You can also enter the names of two big cities, like “Berlin Stockholm”. Google will present you results that relate to travelling primarily by plane, and not, for example, a comparison of the liveability of these cities. Google will highlight various flight options in its built-in flight comparison, while to find train tickets you need to scroll down further. </p>
<p>These results are by no means predefined but instead a result of algorithmic curation. Even without personalisation, search result lists are uniquely created from specific content optimised for specific searches, search engine algorithms, and a user’s query and location. </p>
<p>You can try this yourself with these and other cities. But note that the use of quotation marks, the order of the cities or local versus English spellings can make a difference, since so many companies are trying to optimise for specific searches</p>
<p>Any reader familiar with Google search knows that those alternative results we outlined <a href="https://support.google.com/websearch/answer/134479?hl=en">require other queries</a>. Such queries would need to explicitly establish that the search is about something other than buying clothes or flights. For example by querying “summer clothes colours” or “Berlin Stockholm liveable”. </p>
<p>In any case, the default options that the algorithms select and curate shape what we think as the default. If we are not careful and reflective about our own aims when searching, it will also affect at least some people’s actions. And these actions have very real environmental implications.</p>
<h2>Environmental damage as algorithmic harm</h2>
<p>We suggest calling these environmental implications “algorithmically embodied emissions”. By this, we mean emissions potentially contained in the content that algorithmic information systems – such as search engines or a Facebook or TikTok feed – suggest as their default option. </p>
<p>Our work so far is conceptual though we hope to develop a way to quantify the concept in the future. For now, we can observe that search results tend to suggest high-carbon practices. </p>
<p>And we can note that associated companies like flight comparison services or fast-fashion brands can also optimise their websites for better ranking in search engines. These companies tend to have larger budgets than their more sustainable alternatives (a small organic or repurposed summer clothes brand, for instance).</p>
<p>In recent years, researchers have highlighted the potential harm that algorithmic decision making can do to people, for example by reproducing <a href="https://www.degruyter.com/document/doi/10.18574/nyu/9781479833641.001.0001/html">racial</a> or <a href="https://dl.acm.org/doi/abs/10.1145/3392866">gender</a> <a href="https://www.tandfonline.com/doi/full/10.1080/0960085X.2021.1927212">biases</a>. This is often called <a href="https://journals.sagepub.com/doi/full/10.1177/02683962211010356">algorithmic harm</a>.</p>
<p>The concept of algorithmically embodied emissions asks us to take algorithmic harm even further. It shows that algorithmic decision making has real impacts on both people and planet. </p>
<p>It’s also an example of how algorithmic decision making has higher-order effects beyond the immediate harm done to individuals. In other words: it matters how algorithms function and shape our actions. While the climate crisis accelerates, we have only just started to interrogate how <a href="https://www.slu.se/algorithms">algorithms shape how we think about and act towards the environment</a>. </p>
<hr>
<p><em>In response to this article, a Google spokesperson said:</em></p>
<blockquote>
<p>On Google Search, our goal is to connect people with timely, relevant and helpful information to make the sustainable choice an easier choice. We fundamentally design our Search ranking systems to surface high quality, reliable information on topics like climate change. To complement those efforts, we’ve also developed a number of features to give people helpful context to make informed decisions about sustainability, including helping people quickly access information about the environmental impact of goods and services that they see in results. We collaborate with thousands of partners across multiple industries — from cities and governments to companies and nonprofits — to advance sustainability and climate progress.</p>
</blockquote><img src="https://counter.theconversation.com/content/194334/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Malte Rödl receives funding for this research from Mistra, the Swedish Foundation for Strategic Environmental Research, through the research programme Mistra Environmental Communication.
</span></em></p><p class="fine-print"><em><span>Jutta Haider receives funding from MISTRA, the Swedish foundation for strategic environmental research through the programme MISTRA Environmental Communication. </span></em></p>New research highlights the problem of algorithmically embodied emissions.Malte Rödl, Researcher, Swedish University of Agricultural Sciences and Honorary Research Fellow, University of ManchesterJutta Haider, Professor of Information Studies, University of BoråsLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1790992022-03-27T19:12:14Z2022-03-27T19:12:14ZThere is, in fact, a ‘wrong’ way to use Google. Here are 5 tips to set you on the right path<figure><img src="https://images.theconversation.com/files/454072/original/file-20220324-21-h22dff.jpg?ixlib=rb-1.1.0&rect=23%2C5%2C3902%2C2937&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>I was recently reading comments on a post related to COVID-19, and saw a reply I would classify as misinformation, bordering on conspiracy. I couldn’t help but ask the commenter for evidence. </p>
<p>Their response came with some web links and “do your own research”. I then asked about their research methodology, which turned out to be searching for specific terms on Google. </p>
<p>As an academic, I was intrigued. Academic research aims to establish the truth of a phenomenon based on evidence, analysis and peer review. </p>
<p>On the other hand, a search on Google provides links with content written by known or unknown authors, who may or may not have knowledge in that area, based on a ranking system that either follows the preferences of the user, or the collective popularity of certain sites. </p>
<p>In other words, Google’s algorithms can penalise the truth for not being popular. </p>
<p><a href="https://www.google.com/search/howsearchworks/algorithms">Google Search’s</a> ranking system has a <a href="https://youtu.be/tFq6Q_muwG0">fraction of a second</a> to sort through hundreds of billions of web pages, and index them to find the most relevant and (ideally) useful information. </p>
<p>Somewhere along the way, mistakes get made. And it’ll be a while before these algorithms become foolproof – if ever. Until then, what can you do to make sure you’re not getting the short end of the stick?</p>
<h2>One question, millions of answers</h2>
<p>There are around <a href="https://morningscore.io/how-does-google-rank-websites/">201 known factors</a> on which a website is analysed and ranked by Google’s algorithms. Some of the main ones are:</p>
<ul>
<li>the specific key words used in the search</li>
<li>the meaning of the key words </li>
<li>the relevance of the web page, as assessed by the ranking algorithm<br></li>
<li>the “quality” of the contents </li>
<li>the usability of the web page </li>
<li>and user-specific factors such as their location and profiling data taken from connected Google products, including Gmail, YouTube and Google Maps.</li>
</ul>
<p><a href="https://link.springer.com/article/10.1007/s10676-013-9321-6">Research has shown</a> users pay more attention to higher-ranked results on the first page. And there are known ways to ensure a website makes it to the first page. </p>
<p>One of these is “<a href="https://en.wikipedia.org/wiki/Search_engine_optimization">search engine optimisation</a>”, which can help a web page float into the top results even if its content isn’t necessarily quality. </p>
<p>The other issue is Google Search results <a href="https://mcculloughwebservices.com/2021/01/07/why-google-results-look-different-for-everyone/">are different for different people</a>, sometimes even if they have the exact same search query. </p>
<p>Results are tailored to the user conducting the search. In his book <a href="https://www.penguin.co.uk/books/181/181850/the-filter-bubble/9780241954522.html">The Filter Bubble</a>, Eli Pariser points out the dangers of this – especially when the topic is of a controversial nature. </p>
<p>Personalised search results create alternate versions of the flow of information. Users receive more of what they’ve already engaged with (which is likely also what they already believe). </p>
<p>This leads to a dangerous cycle which can further polarise people’s views, and in which more searching doesn’t necessarily mean getting closer to the truth.</p>
<h2>A work in progress</h2>
<p>While Google Search is a brilliant search engine, it’s also a work in progress. Google is <a href="https://ai.googleblog.com/2020/04/a-scalable-approach-to-reducing-gender.html">continuously addressing various issues</a> related to its performance. </p>
<p>One major challenge relates to societal biases <a href="https://www.kcl.ac.uk/news/artificial-intelligence-is-demonstrating-gender-bias-and-its-our-fault">concerning race and gender</a>. For example, searching Google Images for “truck driver” or “president” returns images of mostly men, whereas “model” and “teacher” returns images of mostly women. </p>
<p>While the results may represent what has <em>historically</em> been true (such as in the case of male presidents), this isn’t always the same as what is <em>currently</em> true – let alone representative of the world we wish to live in.</p>
<p>Some years ago, Google <a href="https://www.theverge.com/2018/1/12/16882408/google-racist-gorillas-photo-recognition-algorithm-ai">reportedly</a> had to block its image recognition algorithms from identifying “gorillas”, after they began classifying images of black people with the term. </p>
<p>Another issue highlighted by health practitioners relates to people <a href="https://www.healthline.com/health/please-stop-using-doctor-google-dangerous">self diagnosing based on symptoms</a>. It’s estimated about <a href="https://onlinelibrary.wiley.com/doi/full/10.5694/mja2.50600">40% of Australians</a> search online for self diagnoses, and there are about 70,000 health-related searches conducted on Google each minute. </p>
<p>There can be serious repercussions for those who <a href="https://www.medicaldirector.com/press/new-study-reveals-the-worrying-impact-of-doctor-google-in-australia">incorrectly interpret</a> information found through “<a href="https://www.ideas.org.au/blogs/dr-google-should-you-trust-it.html">Dr Google</a>” – not to mention what this means in the midst of a pandemic. </p>
<p>Google has delivered a plethora of COVID misinformation related to unregistered medicines, fake cures, mask effectiveness, contact tracing, lockdowns and, of course, vaccines. </p>
<p>According to <a href="https://www.ajtmh.org/view/journals/tpmd/103/4/article-p1621.xml">one study</a>, an estimated 6,000 hospitalisations and 800 deaths during the first few months of the pandemic were attributable to misinformation (specifically the false claim that <a href="https://www.abc.net.au/news/2020-04-28/hundreds-dead-in-iran-after-drinking-methanol-to-cure-virus/12192582">drinking methanol can cure COVID</a>).</p>
<p>To combat this, <a href="https://misinforeview.hks.harvard.edu/article/how-search-engines-disseminate-information-about-covid-19-and-why-they-should-do-better/">Google eventually prioritised</a> authoritative sources in its search results. But there’s only so much Google can do. </p>
<p>We each have a responsibility to make sure we’re thinking critically about the information we come across. What can you do to make sure you’re asking Google the best question for the answer you need?</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/is-google-getting-worse-increased-advertising-and-algorithm-changes-may-make-it-harder-to-find-what-youre-looking-for-166966">Is Google getting worse? Increased advertising and algorithm changes may make it harder to find what you're looking for</a>
</strong>
</em>
</p>
<hr>
<h2>How to Google smarter</h2>
<p>In summary, a Google Search user must be aware of the following facts:</p>
<ol>
<li><p>Google Search will bring you the top-ranked web pages which are also the most relevant to your search terms. Your results will be as good as your terms, so always consider context and how the inclusion of certain terms might affect the result. </p></li>
<li><p>You’re better off starting with a <a href="https://support.google.com/websearch/answer/134479?hl=enr">simple search</a>, and adding more descriptive terms later. For instance, which of the following do you think is a more effective question: “<em>will hydroxychloroquine help cure my COVID?</em>” or “<em>what is hydroxychloroquine used for?</em>”</p></li>
<li><p>Quality content comes from verified (or verifiable) sources. While scouring through results, look at the individual URLs and think about whether that source holds much authority (for instance, is it a government website?). Continue this process once you’re in the page, too, always checking for author credentials and information sources.</p></li>
<li><p>Google may personalise your results based on your previous search history, current location and interests (gleaned through other products such as Gmail, YouTube or Maps). You can use <a href="https://support.google.com/chrome/answer/95464?hl=en&co=GENIE.Platform%3DDesktop">incognito mode</a> to prevent these factors from impacting your search results.</p></li>
<li><p>Google Search isn’t the only option. And you don’t just have to leave your reading to the discretion of its algorithms. There are several other search engines available, including <a href="https://www.bing.com/">Bing</a>, <a href="https://au.yahoo.com/">Yahoo</a>, <a href="https://www.baidu.com/">Baidu</a>, <a href="https://duckduckgo.com/">DuckDuckGo</a> and <a href="https://www.ecosia.org/">Ecosia</a>. Sometimes it’s good to triangulate your results from outside the filter bubble. </p></li>
</ol>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/australias-competition-watchdog-says-google-has-a-monopoly-on-online-advertising-but-how-does-it-work-168939">Australia's competition watchdog says Google has a monopoly on online advertising — but how does it work?</a>
</strong>
</em>
</p>
<hr>
<img src="https://counter.theconversation.com/content/179099/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Muneera Bano does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>There are around 201 known factors on which a website is analysed and ranked by Google’s algorithms.Muneera Bano, Senior Lecturer, Software Engineering, Deakin UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1713372021-12-03T13:43:45Z2021-12-03T13:43:45ZPregnancy apps and online spaces fail to support individuals grieving a pregnancy loss – here’s what to do about it<figure><img src="https://images.theconversation.com/files/433780/original/file-20211124-23-z246zg.jpg?ixlib=rb-1.1.0&rect=44%2C713%2C7395%2C4238&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Social technologies perpetuate a single idea of what constitutes a pregnancy.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/social-distancing-during-covid-19-pandemic-royalty-free-image/1214516643">martin-dm/E+ via Getty Images</a></span></figcaption></figure><p>Approximately <a href="https://doi.org/10.1093/epirev/mxf006">1 in 4 pregnancies in the United States ends in loss</a>. Pregnancy loss, also referred to as miscarriage, is a common reproductive health complication. </p>
<p>Many experience this loss as a significant life event, with a “before” and an “after.” It can cause <a href="https://doi.org/10.1016/0002-9378(92)91839-3">depression and post-traumatic stress disorder</a>. Yet <a href="https://doi.org/10.2307/3178398">society largely stigmatizes and dismisses it</a> by not treating it as a loss that deserves to be grieved.</p>
<p>I <a href="https://scholar.google.com/citations?hl=en&user=ju-VqbUAAAAJ">research the social implications of technology</a>. For the past several years, I’ve been investigating the intersection of pregnancy loss and social technologies. Search engines, social media, online support groups and pregnancy and fertility tracking apps are some of the technologies people use to <a href="https://doi.org/10.1016/j.wombi.2016.01.008">manage pregnancies, share experiences or exchange social support</a>. </p>
<p>My recent research shows these technologies often do not account for pregnancy loss and, as a result, can cause re-traumatization and distress.</p>
<h2>Harmful designs and algorithms</h2>
<p>In a recent study, I conducted <a href="https://doi.org/10.1145/3449201">in-depth interviews with women in the U.S.</a> who had recently experienced pregnancy loss. I found that pregnancy tracking applications failed miserably in considering pregnancy losses. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Woman faces her laptop, head in hands." src="https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=565&fit=crop&dpr=1 754w, https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=565&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/433782/original/file-20211124-17-1lpvlbo.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=565&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">‘Oh, please stop.’</span>
<span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/businesswoman-frustrated-by-bad-new-at-office-desk-royalty-free-image/1124417258">Luis Alvarez/DigitalVision via Getty Images</a></span>
</figcaption>
</figure>
<p>A participant told me, “There’s no way to tell your app, ‘I had a miscarriage. Please stop sending me these updates,’ like, ‘This week, your baby’s the size of a banana or whatever.’ There’s no way to stop those.”</p>
<p>Similarly, advertising algorithms assumed all pregnancies lead to the birth of an alive and healthy baby. Another participant told me, “I was getting ads for maternity clothes. I was just like, ‘Oh, please stop.’”</p>
<p>The design of mobile apps tells a similar story. I conducted <a href="https://doi.org/10.1177/1461444820984473">an analysis of 166 pregnancy-related apps</a> and found 72% do not account for pregnancy loss at all, 18% offer an option to report a loss without providing any support, and the remaining 10% passively link to outside sources.</p>
<p>Another tool people use during pregnancy and loss journeys are online support groups. While groups dedicated to loss can be sources of social support where people may find emotional validation, connect with others and feel seen and less alone, I found they can also foster invalidating and harmful experiences. </p>
<p>One participant reported seeing questions “like ‘Can you eat this certain thing while pregnant?’ You get some people who say, ‘Yes, I ate that all through pregnancy.’ Then you get some people who say, ‘I can’t believe you’re doing that to your body, that’s harmful for you.’” </p>
<p>Overall, the design features and algorithms that underpin content and interactions do real harm by perpetuating a single idea of what constitutes a pregnancy – one that is smooth and leads to a happy ending. By not accounting for pregnancy loss, I contend <a href="https://doi.org/10.1177/1461444820984473">they contribute to its further stigmatization</a>. </p>
<p>My work shows how technology design reinforces stereotypes about experiences like pregnancy loss – and sustains social inequities like marginalization and stigmatization. This, in turn, makes it hard for those experiencing loss to find the resources and support they need.</p>
<h2>A more humane approach</h2>
<p>If you are someone who has experienced pregnancy loss, I am sorry for your loss. Please know that you are not alone. I hope this article helps validate and make visible some of your frustrating experiences.</p>
<p>If you know someone who has experienced a pregnancy loss, know that the harms and challenges I described above are only some of the frustrations they may face. Acknowledge their loss. Ask how you may be able to support them. Get them meals, offer to pet sit or babysit for them, listen to them, sit in their sorrow with them. Know that holidays and anniversaries tend to be tough. Do not say “you will get pregnant again.” Finally, remember that lesbian, gay, bisexual, transgender and queer people also experience pregnancy and loss. </p>
<p>If you are a designer, developer or someone who makes decisions about products and advertising algorithms, I hope this research illustrates some of the real harms users may experience as a result of using products to manage intimate personal experiences like pregnancies. Please consider designing products that consider the full range of pregnancy and other human experiences. Remember that considering pregnancy loss as an outcome does not mean finding other ways to profit from your users’ loss and grief.</p>
<p>[<em>Over 140,000 readers rely on The Conversation’s newsletters to understand the world.</em> <a href="https://memberservices.theconversation.com/newsletters/?source=inline-140ksignup">Sign up today</a>.]</p><img src="https://counter.theconversation.com/content/171337/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Nazanin Andalibi does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>New research shows technologies like pregnancy apps do not account for pregnancy loss 72% of the time, causing real harm to users.Nazanin Andalibi, Assistant Professor of Information, University of MichiganLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1453752021-09-10T12:47:06Z2021-09-10T12:47:06ZHow ‘engagement’ makes you vulnerable to manipulation and misinformation on social media<figure><img src="https://images.theconversation.com/files/420340/original/file-20210909-13-bbylok.jpg?ixlib=rb-1.1.0&rect=0%2C7%2C4802%2C3184&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">People tend to view social media posts more favorably when more people have liked, commented on or shared them, regardless of the quality of the posts.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/news-photo/this-picture-taken-on-december-18-2018-shows-myanmar-youths-news-photo/1074380170">Sai Aung Main/AFP via Getty Images</a></span></figcaption></figure><p><em>An updated version of this article was published on Sept. 20, 2021. <a href="https://theconversation.com/facebooks-algorithms-fueled-massive-foreign-propaganda-campaigns-during-the-2020-election-heres-how-algorithms-can-manipulate-you-168229">Read it here</a>.</em></p>
<p>Facebook has been <a href="https://about.fb.com/news/2021/02/reducing-political-content-in-news-feed/">quietly experimenting</a> with reducing the amount of political content it puts in users’ news feeds. The move is a tacit acknowledgment that the way the company’s algorithms work <a href="https://www.wired.com/story/facebook-quietly-makes-big-admission-political-content/">can be a problem</a>.</p>
<p>The heart of the matter is the distinction between provoking a response and providing content people want. Social media algorithms – the rules their computers follow in deciding the content that you see – rely heavily on people’s behavior to make these decisions. In particular, they watch for content that people respond to or “engage” with by liking, commenting and sharing.</p>
<p>As a <a href="https://scholar.google.com/citations?user=f_kGJwkAAAAJ&hl=en">computer scientist</a> who studies the ways large numbers of people interact using technology, I understand the logic of using the <a href="https://www.penguinrandomhouse.com/books/175380/the-wisdom-of-crowds-by-james-surowiecki/">wisdom of the crowds</a> in these algorithms. I also see substantial pitfalls in how the social media companies do so in practice.</p>
<h2>From lions on the savanna to likes on Facebook</h2>
<p>The concept of the wisdom of crowds assumes that using signals from others’ actions, opinions and preferences as a guide will lead to sound decisions. For example, <a href="https://www.investopedia.com/terms/p/prediction-market.asp">collective predictions</a> are normally more accurate than individual ones. Collective intelligence is used to predict <a href="https://augur.net/">financial markets, sports</a>, <a href="https://iemweb.biz.uiowa.edu/">elections</a> and even <a href="https://www.centerforhealthsecurity.org/our-work/Center-projects/disease-prediction-project.html">disease outbreaks</a>. </p>
<p>Throughout millions of years of evolution, these principles have been coded into the human brain in the form of cognitive biases that come with names like <a href="https://doi.org/10.1086/208859">familiarity</a>, <a href="http://socialpsychonline.com/2016/03/the-mere-exposure-effect/">mere-exposure</a> and <a href="https://www.psychologytoday.com/us/blog/stronger-the-broken-places/201708/the-bandwagon-effect">bandwagon effect</a>. If everyone starts running, you should also start running; maybe someone saw a lion coming and running could save your life. You may not know why, but it’s wiser to ask questions later. </p>
<p>Your brain picks up clues from the environment – including your peers – and uses <a href="https://global.oup.com/academic/product/simple-heuristics-that-make-us-smart-9780195143812">simple rules</a> to quickly translate those signals into decisions: Go with the winner, follow the majority, copy your neighbor. These rules work remarkably well in typical situations because they are based on sound assumptions. For example, they assume that people often act rationally, it is unlikely that many are wrong, the past predicts the future, and so on.</p>
<p>Technology allows people to access signals from much larger numbers of other people, most of whom they do not know. Artificial intelligence applications make heavy use of these popularity or “engagement” signals, from selecting search engine results to recommending music and videos, and from suggesting friends to ranking posts on news feeds. </p>
<h2>Not everything viral deserves to be</h2>
<p>Our research shows that virtually all web technology platforms, such as social media and news recommendation systems, have a strong <a href="http://doi.org/10.1002/asi.24121">popularity bias</a>. When applications are driven by cues like engagement rather than explicit search engine queries, popularity bias can lead to harmful unintended consequences. </p>
<p>Social media like Facebook, Instagram, Twitter, YouTube and TikTok rely heavily on AI algorithms to rank and recommend content. These algorithms take as input what you “like,” comment on and share – in other words, content you engage with. The goal of the algorithms is to maximize engagement by finding out what people like and ranking it at the top of their feeds. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/doWZHFnVPQ8?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">A primer on the Facebook algorithm.</span></figcaption>
</figure>
<p>On the surface this seems reasonable. If people like credible news, expert opinions and fun videos, these algorithms should identify such high-quality content. But the wisdom of the crowds makes a key assumption here: that recommending what is popular will help high-quality content “bubble up.” </p>
<p>We <a href="http://doi.org/10.1038/s41598-018-34203-2">tested this assumption</a> by studying an algorithm that ranks items using a mix of quality and popularity. We found that in general, popularity bias is more likely to lower the overall quality of content. The reason is that engagement is not a reliable indicator of quality when few people have been exposed to an item. In these cases, engagement generates a noisy signal, and the algorithm is likely to amplify this initial noise. Once the popularity of a low-quality item is large enough, it will keep getting amplified. </p>
<p>Algorithms aren’t the only thing affected by engagement bias – it can <a href="https://www.scientificamerican.com/article/information-overload-helps-fake-news-spread-and-social-media-knows-it/">affect people</a>, too. Evidence shows that information is transmitted via “<a href="https://doi.org/10.1371/journal.pone.0184148">complex contagion</a>,” meaning the more times someone is exposed to an idea online, the more likely they are to adopt and reshare it. When social media tells people an item is going viral, their cognitive biases kick in and translate into the irresistible urge to pay attention to it and share it.</p>
<h2>Not-so-wise crowds</h2>
<p>We recently ran an experiment using <a href="https://fakey.iuni.iu.edu/">a news literacy app called Fakey</a>. It is a game developed by our lab, which simulates a news feed like those of Facebook and Twitter. Players see a mix of current articles from fake news, junk science, hyper-partisan and conspiratorial sources, as well as mainstream sources. They get points for sharing or liking news from reliable sources and for flagging low-credibility articles for fact-checking. </p>
<p>We found that players are <a href="https://doi.org/10.37016/mr-2020-033">more likely to like or share and less likely to flag</a> articles from low-credibility sources when players can see that many other users have engaged with those articles. Exposure to the engagement metrics thus creates a vulnerability.</p>
<p><iframe id="HoqGE" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/HoqGE/1/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>The wisdom of the crowds fails because it is built on the false assumption that the crowd is made up of diverse, independent sources. There may be several reasons this is not the case. </p>
<p>First, because of people’s tendency to associate with similar people, their online neighborhoods are not very diverse. The ease with which a social media user can unfriend those with whom they disagree pushes people into homogeneous communities, often referred to as <a href="https://doi.org/10.1007/s42001-020-00084-7">echo chambers</a>. </p>
<p>Second, because many people’s friends are friends of each other, they influence each other. A <a href="https://doi.org/10.1126/science.1121066">famous experiment</a> demonstrated that knowing what music your friends like affects your own stated preferences. Your social desire to conform distorts your independent judgment. </p>
<p>Third, popularity signals can be gamed. Over the years, search engines have developed sophisticated techniques to counter so-called “<a href="https://www.webopedia.com/TERM/L/link_farming.html">link farms</a>” and other schemes to manipulate search algorithms. Social media platforms, on the other hand, are just beginning to learn about their own <a href="https://theconversation.com/misinformation-on-social-media-can-technology-save-us-69264">vulnerabilities</a>. </p>
<p>People aiming to manipulate the information market have created <a href="https://www.washingtonpost.com/technology/2020/10/13/black-fake-twitter-accounts-for-trump/">fake accounts</a>, like trolls and <a href="https://cacm.acm.org/magazines/2016/7/204021-the-rise-of-social-bots/fulltext">social bots</a>, and <a href="https://ojs.aaai.org/index.php/ICWSM/article/view/18075">organized</a> <a href="https://www.washingtonpost.com/politics/turning-point-teens-disinformation-trump/2020/09/15/c84091ae-f20a-11ea-b796-2dd09962649c_story.html">fake networks</a>. They have <a href="http://doi.org/10.1038/s41467-018-06930-7">flooded the network</a> to create the appearance that a <a href="https://www.newsweek.com/2020/10/23/qanon-conspiracy-theories-draw-new-believers-scientists-take-aim-misinformation-pandemic-1538901.html">conspiracy theory</a> or a <a href="https://ojs.aaai.org/index.php/ICWSM/article/view/14127">political candidate</a> is popular, tricking both platform algorithms and people’s cognitive biases at once. They have even <a href="https://doi.org/10.1038/s41586-019-1507-6">altered the structure of social networks</a> to create <a href="https://doi.org/10.1371/journal.pone.0147617">illusions about majority opinions</a>. </p>
<p>[<em>Over 110,000 readers rely on The Conversation’s newsletter to understand the world.</em> <a href="https://theconversation.com/us/newsletters/the-daily-3?utm_source=TCUS&utm_medium=inline-link&utm_campaign=newsletter-text&utm_content=100Ksignup">Sign up today</a>.]</p>
<h2>Dialing down engagement</h2>
<p>What to do? Technology platforms are currently on the defensive. They are becoming more <a href="https://www.nytimes.com/2020/10/09/technology/twitter-election-ban-features.html">aggressive</a> during elections in <a href="https://www.socialmediatoday.com/news/facebook-outlines-its-evolving-efforts-to-combat-misinformation-ahead-of-ne/597129/">taking down fake accounts and harmful misinformation</a>. But these efforts can be akin to a game of <a href="https://www.marketplace.org/shows/marketplace-tech/facebook-plays-whack-a-mole-with-foreign-election-interference/">whack-a-mole</a>. </p>
<p>A different, preventive approach would be to add <a href="https://www.theguardian.com/commentisfree/2020/jun/29/social-distancing-social-media-facebook-misinformation">friction</a>. In other words, to slow down the process of spreading information. High-frequency behaviors such as automated liking and sharing could be inhibited by <a href="https://www.cloudflare.com/learning/bots/how-captchas-work/">CAPTCHA</a> tests or fees. This would not only decrease opportunities for manipulation, but with less information people would be able to pay more attention to what they see. It would leave less room for engagement bias to affect people’s decisions.</p>
<p>It would also help if social media companies adjusted their algorithms to rely less on engagement to determine the content they serve you.</p><img src="https://counter.theconversation.com/content/145375/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Filippo Menczer receives funding from Knight Foundation, Craig Newmark Philanthropies, DARPA and AFOSR. </span></em></p>You have evolved to tap into the wisdom of the crowds. But on social media your cognitive biases can lead you astray.Filippo Menczer, Luddy Distinguished Professor of Informatics and Computer Science, Indiana UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1669662021-09-01T06:39:22Z2021-09-01T06:39:22ZIs Google getting worse? Increased advertising and algorithm changes may make it harder to find what you’re looking for<figure><img src="https://images.theconversation.com/files/418822/original/file-20210901-13-1nyiv7m.jpg?ixlib=rb-1.1.0&rect=5%2C8%2C1991%2C1320&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>Over the past 25 years, the name “Google” has become synonymous with the idea of searching for anything online. In much the same way “to Hoover” means to use a vacuum cleaner, <a href="https://www.wired.com/story/just-google-it-a-short-history-of-a-newfound-verb/">dictionaries have recognised “to Google”</a> as meaning to undertake an online search using any available service.</p>
<p>Former competitors such as <a href="https://www.engadget.com/2018-09-25-google-ask-jeeves-altavista-dogpile-dead-search-engines.html">AltaVista and AskJeeves</a> are long dead, and existing alternatives such as Bing and DuckDuckGo currently pose little threat to Google’s dominance. But shifting our web searching habits to a single supplier has significant risks.</p>
<p>Google also dominates in the web browser market (<a href="https://gs.statcounter.com/">almost two-thirds of browsers are Chrome</a>) and web advertising (Google Ads has an <a href="https://www.cnbc.com/2021/05/18/how-does-google-make-money-advertising-business-breakdown-.html">estimated 29% share</a> of all digital advertising in 2021). This combination of browser, search and advertising has drawn considerable interest from <a href="https://clearcode.cc/blog/antitrust-investigations-gafa/">competition and antitrust regulators</a> around the world.</p>
<p>Leaving aside the commercial interests, is Google actually delivering when we <em>Google</em>? Are the search results (which clearly influence the content we consume) giving us the answers we want?</p>
<h2>Advertising giant</h2>
<p>More than 80% of Alphabet’s revenue comes from <a href="https://www.cnbc.com/2021/05/18/how-does-google-make-money-advertising-business-breakdown-.html">Google advertising</a>. At the same time, around <a href="https://www.statista.com/statistics/216573/worldwide-market-share-of-search-engines/">85% of the world’s search engine activity</a> goes through Google.</p>
<p>Clearly there is significant commercial advantage in selling advertising while at the same time controlling the results of most web searches undertaken around the globe.</p>
<p>This can be seen clearly in search results. <a href="https://www.nngroup.com/articles/scrolling-and-attention/">Studies have shown</a> internet users are less and less prepared to scroll down the page or spend less time on content below the “fold” (the limit of content on your screen). This makes the space at the top of the search results more and more valuable.</p>
<p>In the example below, you might have to scroll three screens down before you find actual search results rather than paid promotions.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=1120&fit=crop&dpr=1 600w, https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=1120&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=1120&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=1407&fit=crop&dpr=1 754w, https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=1407&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/418608/original/file-20210831-19-4k0l1d.PNG?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=1407&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">In a simple Google search (for ‘buy shoes’), you have to scroll a long way to find the results.</span>
<span class="attribution"><span class="source">Author provided</span></span>
</figcaption>
</figure>
<p>While Google (and indeed many users) might argue that the results are still helpful and save time, it’s clear the design of the page and the prominence given to paid adverts will influence behaviour. All of this is reinforced by the use of a <a href="https://ads.google.com/home/campaigns/search-ads/">pay-per-click</a> advertising model which is founded on enticing users to click on adverts.</p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"757606109322633216"}"></div></p>
<h2>Annoyance</h2>
<p>Google’s influence expands beyond web search results. More than <a href="https://blog.youtube/press/">2 billion people</a> use the Google-owned YouTube each month (just counting logged-in users), and it is often considered the number one platform for online advertising. </p>
<p>Although YouTube is as ubiquitous to video-sharing as Google is to search, YouTube users have an option to avoid ads: paying for a premium subscription. However, only a minuscule fraction of users take the paid option.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/rvCq1JeMgh8?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">Why are there so many ads on YouTube lately?</span></figcaption>
</figure>
<h2>Evolving needs</h2>
<p>The complexity (and expectations) of search engines has increased over their lifetime, in line with our dependence on technology. </p>
<p>For example, someone trying to explore a tourist destination may be tempted to search “What should I do to visit the <a href="https://northernterritory.com/alice-springs-and-surrounds/see-and-do/simpsons-gap">Simpsons Gap</a>”. </p>
<p>The Google search result will show a number of results, but from the user perspective the information is distributed across multiple sites. To obtain the desired information users need to visit a number of websites. </p>
<p>Google is working on bringing this information together. The search engine now uses sophisticated “natural language processing” software called <a href="https://blog.google/products/search/search-language-understanding-bert/">BERT</a>, developed in 2018, that tries to identify the <em>intention</em> behind a search, rather than simply searching strings of text. <a href="https://www.wired.com/1997/06/jeeves-finds-the-way/">AskJeeves tried something similar</a> in 1997, but the technology is now more advanced.</p>
<p>BERT will soon be succeeded by MUM (<a href="https://blog.google/products/search/introducing-mum/">Multitask Unified Model</a>), which tries to go a step further and understand the context of a search and provide more refined answers. Google claims MUM may be 1000 times more powerful than BERT, and be able to provide the kind of advice a human expert might for questions without a direct answer.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/s7t4lLgINyo?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">Google MUM MultiTask Unified Model Introduction.</span></figcaption>
</figure>
<h2>Are we now locked into Google?</h2>
<p>Given the market share and influence Google has in our daily lives, it might seem impossible to think of alternatives. However, Google is not the only show in town. Microsoft’s Bing search engine has a modest level of popularity in the United States, although it will struggle to escape the Microsoft brand. </p>
<p>Another option that claims to be free from ads and ensure user privacy, <a href="https://duckduckgo.com/">DuckDuckGo</a>, has seen a growing level of interest - perhaps helped through association with the <a href="https://techcrunch.com/2016/05/31/tor-switches-to-duckduckgo-search-results-by-default/">TOR browser project</a>. </p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1429744930625409033"}"></div></p>
<p>While Google may be dominating with its search engine service, it also covers <a href="https://ai.google/">artificial intelligence</a>, <a href="https://health.google/">healthcare</a>, <a href="https://waymo.com/">autonomous vehicles</a>, <a href="https://cloud.google.com/">cloud computing services</a>, <a href="https://www.google.com/intl/en_au/chromebook/">computing devices</a> and a plethora of <a href="https://store.google.com/category/connected_home?hl=en-GB">home automation devices</a>. Even if we can move away from Google’s grasp in our web browsing activities, there is a whole new range of future challenges for consumers on the horizon.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/robot-take-the-wheel-waymo-has-launched-a-self-driving-taxi-service-147908">Robot take the wheel: Waymo has launched a self-driving taxi service</a>
</strong>
</em>
</p>
<hr>
<img src="https://counter.theconversation.com/content/166966/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.</span></em></p>More advertising and “smarter” search algorithms are changing how Googling works.Mohiuddin Ahmed, Lecturer of Computing & Security, Edith Cowan UniversityPaul Haskell-Dowland, Associate Dean (Computing and Security), Edith Cowan UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1604922021-05-24T12:09:02Z2021-05-24T12:09:02ZWhy do I need anything other than Google to answer a question?<figure><img src="https://images.theconversation.com/files/400406/original/file-20210512-15-2qezsd.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C4043%2C1997&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Scholars can be more reliable than search engines.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/news-photo/silhouettes-of-people-holding-laptops-are-seen-in-front-of-news-photo/1026614170">Anadolu Agency/Getty Images</a></span></figcaption></figure><figure class="align-left ">
<img alt="" src="https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=293&fit=crop&dpr=1 600w, https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=293&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=293&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=368&fit=crop&dpr=1 754w, https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=368&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/281719/original/file-20190628-76743-26slbc.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=368&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption"></span>
</figcaption>
</figure>
<p><em><a href="https://theconversation.com/us/topics/curious-kids-us-74795">Curious Kids</a> is a series for children of all ages. If you have a question you’d like an expert to answer, send it to <a href="mailto:curiouskidsus@theconversation.com">curiouskidsus@theconversation.com</a>.</em></p>
<hr>
<blockquote>
<p><strong>Why do I need a scholar to answer a question if there is Google? – Harrison F., age 13, Brookline, Massachusetts</strong></p>
</blockquote>
<hr>
<p>Imagine you’re researching something. Whether you’re a fourth grader who needs to find out how volcanoes erupt or you’re an adult looking for more information regarding a news article, you might want to quickly look something up on the internet. What could go wrong?</p>
<p>Google’s search engine may seem to have all the answers to your questions. But where does that information come from? Who selects the websites that display when you enter “volcanic eruption” in the search box? Who decides which item shows up first and in what order the rest will follow?</p>
<p>I think about these questions a lot because of what I do for a living: helping University of Memphis scholars <a href="https://scholar.google.com/citations?user=dFsRzLUAAAAJ&hl=en&oi=ao">communicate about their work</a> with academic peers and the public. </p>
<p>These scholars are experts who have worked and studied for a long time to learn all they can about a topic. They answer questions by combining their knowledge with the <a href="https://www.sciencebuddies.org/science-fair-projects/science-fair/steps-of-the-scientific-method">scientific method</a> to discover new things. </p>
<h2>Page, Brin and PageRank</h2>
<p>When <a href="https://www.theverge.com/2019/12/4/20994361/google-alphabet-larry-page-sergey-brin-sundar-pichai-co-founders-ceo-timeline">Larry Page and Sergey Brin</a> created Google’s search engine in 1996 as Stanford University computer science students, they were trying to establish a fast way to easily find things on the internet. At the time, <a href="https://www.iflscience.com/technology/this-website-simulates-the-pain-of-loading-the-internet-in-the-90s/">searching through the web was slow and difficult</a>, making it hard to find the best information.</p>
<p>They invented an algorithm, a detailed step-by-step instruction set or formula, called <a href="https://searchengineland.com/what-is-google-pagerank-a-guide-for-searchers-webmasters-11068">PageRank</a>. It works by estimating the quality of a webpage by measuring the number and quality of other pages that link to it. When you search on Google, its search engine returns the highest ranked pages related to what you’re looking for. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/IKXvSKaI2Ko?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">PageRank, explained.</span></figcaption>
</figure>
<h2>Some drawbacks</h2>
<p>Googling became so fast it can seem instantaneous.</p>
<p>But the results you see when you do a Google search can be influenced by other things besides PageRank, including whether <a href="https://www.nature.com/articles/s41599-017-0021-4">advertisers are paying Google</a> to make their websites show up higher than they otherwise might. <a href="https://theconversation.com/its-not-just-a-social-media-problem-how-search-engines-spread-misinformation-152155">Google’s algorithms</a> factor in hundreds of other variables, including what sites you’ve clicked on in the past and how recently a page was updated.</p>
<p>Unlike scholars, Google’s search engine can’t automatically decide which sources are the most important, most accurate or most significant. That is, Google searches don’t necessarily identify <a href="https://theconversation.com/its-not-just-a-social-media-problem-how-search-engines-spread-misinformation-152155">objective and reliable</a> information.</p>
<p>You may consider switching to another search engine like Microsoft’s Bing or one that specifically promotes the privacy of your information like DuckDuckGo. But many of these alternatives have the same shortcomings. </p>
<h2>How scholars communicate</h2>
<p>Scholars often communicate by publishing research papers. Each paper emphasizes a single idea that adds something to a discussion. It may be the new result from an experiment or a new observation. Other scholars then read that paper and discuss it.</p>
<p>Knowledgeable people can take stock of the same set of facts and still have different perspectives, which means there isn’t necessarily one right answer to a question. Over time this back and forth leads to some generally accepted principles and concepts.</p>
<p>This cycle of research, review and discussion has been <a href="https://www.historyofinformation.com/detail.php?entryid=2661">around since the first</a> <a href="https://doi.org/10.1098/rstb.2014.0380">academic journals were published in 1665</a>. As new discoveries are made, ideas can change.</p>
<p>One way researchers show what other ideas they consider in their work is through scholarly citations. You’ve surely seen them before – they are in the reference section at the back of nonfiction books or at the bottom of Wikipedia articles. Each points to another work.</p>
<p>These citations tell you what other books and sources the author of what you’re reading considered – and how they came to form the ideas. If multiple scholars <a href="https://doi.org/10.1023/b:scie.0000027310.68393.bc">use the same ideas as building blocks</a> for their own concepts, and then their ideas, in turn, are used as building blocks for other ideas, it continually leads to a cycle of innovation. </p>
<p>This discovery process isn’t influenced by advertisers – even if it can be partially shaped by whether or not scholars <a href="https://theconversation.com/when-big-companies-fund-academic-research-the-truth-often-comes-last-119164">can get funding</a> to pursue a particular kind of research. </p>
<p>Many of the ideas you find on the internet originate from scholarship but are vulnerable to bias and advertising pressure in a way most scholars are not. We need scholars because they provide a complete picture, the most up-to-date information, derived from their wisdom and deeply considered perspective.</p>
<p>The internet makes locating information easier than at any other point in human history. But as <a href="https://www.relicsworld.com/albert-einstein/information-is-not-knowledge-the-only-source-of-knowledge-is-experience-you-author-albert-einstein">Albert Einstein</a> said, “Information is not knowledge. The only source of knowledge is experience.”</p>
<hr>
<p><em>Hello, curious kids! Do you have a question you’d like an expert to answer? Ask an adult to send your question to <a href="mailto:curiouskidsus@theconversation.com">CuriousKidsUS@theconversation.com</a>. Please tell us your name, age and the city where you live.</em></p>
<p><em>And since curiosity has no age limit – adults, let us know what you’re wondering, too. We won’t be able to answer every question, but we will do our best.</em></p><img src="https://counter.theconversation.com/content/160492/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Cody Behles does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Unlike scholars, Google’s search engine can’t automatically decide which sources are the most important, most accurate or most significant.Cody Behles, Director of Innovation & Research Support, University of MemphisLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1575162021-03-24T12:19:12Z2021-03-24T12:19:12ZFuture of journalism: study explains why some news stories get more clicks from social media than others<figure><img src="https://images.theconversation.com/files/391142/original/file-20210323-21-dxuy4.jpg?ixlib=rb-1.1.0&rect=0%2C8%2C5444%2C3586&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">To click or not to click?</span> <span class="attribution"><span class="source">Hadrian via Shutterstock</span></span></figcaption></figure><p>Few industries have suffered more disruption from the internet than the news media. Over two decades or more, journalism has been hit by a “<a href="https://www.thedrum.com/opinion/2018/02/16/the-perfect-storm-thats-blowing-up-media">perfect storm</a>” due to the loss of geographical monopolies that national and regional news organisations once enjoyed as well as the emergence of amateur content producers such as bloggers.</p>
<p>Perhaps what is the most disruptive element of this assault on news organisations is the “<a href="https://www.researchgate.net/publication/346394853_The_Unbundling_of_Journalism">unbundling</a>” of content production and content aggregation – or, in layperson’s terms, the fact that most people aren’t getting their news directly from news sites but via social media or other places on the internet. </p>
<p>This has led to more than a decade of argument between media companies and news aggregators such as Apple News, Google and Facebook. News aggregators tend to post headlines and short extracts of articles, linking through to the site on which they were originally published. Every click brings <a href="https://www.journalism.org/2014/03/13/social-search-direct/">additional traffic</a> to the news producer’s site – and the all-important advertising revenue.</p>
<p>Media tycoon Rupert Murdoch has been particularly outspoken on news aggregators, referring to them as <a href="https://www.mediaite.com/online/rupert-murdoch-begins-blocking-new-aggregators-search-engines/">“parasites”</a>. His company News Corporation and other media organisations <a href="https://www.theguardian.com/media/greenslade/2009/apr/06/google-wallstreetjournal">have accused them</a> of preventing readers from going to their sites – “stealing” advertising revenues by “free riding” on their content. More than 11 years ago, in an interview with Sky News Australia, the <a href="https://www.theguardian.com/media/2009/nov/09/murdoch-google">news mogul said</a> he would consider removing the content from his news sites from Google’s search index. It never happened.</p>
<p>In February, the Australian government <a href="https://www.bbc.co.uk/news/world-australia-56163550">passed</a> legislation, the first of its kind in the world aiming to make the likes of Apple, Google and Facebook pay for news content. After a short stand-off between Facebook and Australian news organisations, during which Facebook refused to post any Australian news content on subscribers’ newsfeeds, a deal was struck, setting up a <a href="https://www.accc.gov.au/system/files/Revised%20explanatory%20memorandum.pdf">bargaining code</a> for publishers and aggregators.</p>
<h2>Symbiotic relationship</h2>
<p>A central debate on news aggregators is whether they are harmful to news producers by <a href="https://www.tse-fr.eu/sites/default/files/TSE/documents/doc/wp/2018/wp_tse_912.pdf">raking off advertising revenues</a> or beneficial by reducing search time and costs for consumers. Our <a href="https://pubsonline.informs.org/doi/abs/10.1287/mnsc.2015.2237">research examined</a> whether news producers and aggregators can have a beneficial symbiotic relationship. We developed news aggregator apps for the iPhone and iPad and carried out our own field experiments.</p>
<p>The apps aggregated news from 13 major news publishers in Switzerland (with their permission) and were available for download by anyone in that country. The two-week field experiment with the iPhone app involved more than 2,000 users who viewed extracts of almost 5,000 articles a total of more than 32,000 times. The iPad app ran for 16 weeks, during which we had close to 1,400 users viewing extracts of almost 30,000 articles more than 65,000 times.</p>
<p>We varied the amount of text in the extracts, and experimented by accompanying some extracts with an image. We also looked at whether it made a difference if there were a number of other competing articles on the same issue. What we wanted to find out was how likely readers were in various different scenarios to click through to read the full article at the news publisher’s site.</p>
<h2>How the app works</h2>
<p>The default length of the snippet of text in our experiment is 245 characters – which we found was the average number of characters of the snippets in Google News. </p>
<p>We then reduced or increased the number of characters in increments of 20%. The longest snippet we used was 343 characters (+40%) because of the constraint of our copyright agreements with the news providers. The shortest snippet on our iPhone app is 98 characters (-60%). In our iPad app, we sometimes show no snippet at all (only the headline and corresponding image). </p>
<p>We found that as the extracts grew longer, people were less likely to click through to the article on its original site. It appeared that an article’s headline could often provide all the information the audience needed. Any additional information provided by the aggregator, in the form of snippets of text or images, actually decreased click-through rates.</p>
<p>There is massive difference in click-through from iPhone and iPad. The iPad has a richer interface and is the closest to a web browser – which means the limitations of the mobile phone interface may increase the click-through rate. Nevertheless, decline in click-through rates is consistent across both platforms.</p>
<p>Interestingly, the opposite happens when the snippets of several related articles of the same story compete for readers’ attention. Aggregators tend to group these snippets together, which creates direct competition for readers. We found that in cases like this, 30% of readers do not click through to any article and 66% of readers click through to only one article – paradoxically, it’s the snippet with longer text and accompanying images that gets the clicks.</p>
<p>The results of this experiment present a dilemma for news publishers. On the one hand they know that the more information they allow an aggregator to reproduce in terms of text and images, the less likely readers are to visit their site. On the other hand, by limiting the amount of text or images that an aggregator is allowed to reproduce they risk losing out to their competitors who might not follow the same strategy. </p>
<p>Our research leaves us with two insights: one is that news organisations will need to keep experimenting with ways to get people to their sites. The other is that the news industry as a whole needs to negotiate with news aggregators to ensure fair treatment for all.</p><img src="https://counter.theconversation.com/content/157516/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Juliana Sutanto does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>News organisations are wracking their brains on how to get people to their sites.Juliana Sutanto, Professor of Information Systems, Lancaster UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1521552021-03-10T18:51:43Z2021-03-10T18:51:43ZIt’s not just a social media problem – how search engines spread misinformation<figure><img src="https://images.theconversation.com/files/388633/original/file-20210309-19-162kllo.jpg?ixlib=rb-1.1.0&rect=0%2C8%2C6000%2C3979&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Search engines often serve up a distorting blend of information and misinformation.</span> <span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/photo/observe-royalty-free-image/1282737925?adppopup=true">Crispin la valiente/Moment via Getty Images</a></span></figcaption></figure><p>Search engines are one of society’s primary gateways to information and people, but they are also conduits for misinformation. Similar to <a href="https://theconversation.com/hate-cancel-culture-blame-algorithms-129402">problematic social media algorithms</a>, search engines learn to serve you what you and others have clicked on before. Because people are drawn to the sensational, this dance between algorithms and human nature can foster the spread of misinformation.</p>
<p>Search engine companies, like most online services, make money not only by selling ads, but also by tracking users and selling their data <a href="https://www.eff.org/deeplinks/2020/03/google-says-it-doesnt-sell-your-data-heres-how-company-shares-monetizes-and">through real-time bidding</a> on it. People are often led to misinformation by their desire for sensational and entertaining news as well as information that is either controversial or confirms their views. One study found that more popular YouTube videos about diabetes are <a href="https://theconversation.com/biases-in-algorithms-hurt-those-looking-for-information-on-health-140616">less likely to have medically valid information</a> than less popular videos on the subject, for instance.</p>
<p>Ad-driven search engines, like social media platforms, are designed to reward clicking on enticing links because it helps the search companies boost their business metrics. As a researcher who <a href="https://scholar.google.com/citations?user=H4dLAw0AAAAJ&hl=en">studies the search and recommendation systems</a>, I and my colleagues show that this dangerous combination of corporate profit motive and individual susceptibility <a href="http://chiragshah.org/papers/Bias_2021a.pdf">makes the problem difficult to fix</a>.</p>
<h2>How search results go wrong</h2>
<p>When you click on a search result, the search algorithm learns that the link you clicked is relevant for your search query. This is called <a href="https://link.springer.com/referenceworkentry/10.1007%2F978-0-387-39940-9_462">relevance feedback</a>. This feedback helps the search engine give higher weight to that link for that query in the future. If enough people click on that link enough times, thus giving strong relevance feedback, that website starts coming up higher in search results for that and related queries. </p>
<p>People are <a href="https://www.smartinsights.com/search-engine-optimisation-seo/seo-analytics/comparison-of-google-clickthrough-rates-by-position/">more likely to click on links shown up higher</a> on the search results list. This creates a positive feedback loop – the higher a website shows up, the more the clicks, and that in turn makes that website move higher or keep it higher. Search engine optimization techniques use this knowledge to increase the visibility of websites. </p>
<p>There are two aspects to this misinformation problem: how a search algorithm is evaluated and how humans react to headlines, titles and snippets. Search engines, like most online services, are judged using an array of metrics, one of which is user engagement. It is in the search engine companies’ best interest to give you things that you want to read, watch or simply click. Therefore, as a search engine or any recommendation system creates a list of items to present, it calculates the likelihood that you’ll click on the items. </p>
<p>Traditionally, this was meant to bring out the information that would be most relevant. However, the notion of relevance has gotten fuzzy because people have been using search to find <a href="https://www.dummies.com/web-design-development/search-engine-optimization/why-people-use-search-engines-research-shopping-and-entertainment/">entertaining search results as well as truly relevant information</a>.</p>
<p>Imagine you are looking for a piano tuner. If someone shows you a video of a cat playing a piano, would you click on it? Many would, even if that has nothing to do with piano tuning. The search service feels validated with positive relevance feedback and learns that it is OK to show a cat playing a piano when people search for piano tuners. </p>
<p>In fact, it is even better than showing the relevant results in many cases. People like watching funny cat videos, and the search system gets more clicks and user engagement. </p>
<p>This might seem harmless. So what if people get distracted from time to time and click on results that aren’t relevant to the search query? The problem is that people are drawn to exciting images and sensational headlines. They <a href="https://www.aaai.org/ocs/index.php/ICWSM/ICWSM15/paper/viewFile/10568/10535">tend to click on conspiracy theories and sensationalized news</a>, not just cats playing piano, and do so <a href="https://www.buzzfeednews.com/article/craigsilverman/viral-fake-election-news-outperformed-real-news-on-facebook">more than clicking on real news</a> or relevant information.</p>
<h2>Famous but fake spiders</h2>
<p>In 2018, searches for “new deadly spider” <a href="https://trends.google.com/trends/explore?date=today%205-y&geo=US&q=new%20deadly%20spider">spiked on Google</a> following a Facebook post that claimed a new deadly spider killed several people in multiple states. My colleagues and I analyzed the top 100 results from Google search for “new deadly spider” during the first week of this trending query. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Distribution of search results for 'new deadly spider' on Google" src="https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=322&fit=crop&dpr=1 600w, https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=322&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=322&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=405&fit=crop&dpr=1 754w, https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=405&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/386809/original/file-20210227-23-dbonwc.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=405&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">The first two pages of Google search results for ‘new deadly spider’ in August 2018 (shaded area) were related to the original fake news post about that subject, not debunking or otherwise factual information.</span>
<span class="attribution"><span class="source">Chirag Shah</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>It turned out this story <a href="https://www.snopes.com/fact-check/new-deadly-spider/">was fake</a>, but people searching for it were largely exposed to misinformation related to the original fake post. As people continued clicking and sharing that misinformation, Google continued serving those pages at the top of the search results. </p>
<p>This pattern of thrilling and unverified stories emerging and people clicking on them continues, with people apparently either being unconcerned with the truth or believing that if a trusted service such as Google Search is showing these stories to them then the stories must be true. More recently, a <a href="https://www.cnn.com/2020/10/21/politics/coronavirus-lab-theory-yan-bannon-invs/index.html">disproven report</a> claiming China let the coronavirus leak from a lab gained traction on search engines because of this vicious cycle.</p>
<h2>Spot the misinformation</h2>
<p>To test how well people discriminate between accurate information and misinformation, we designed a simple game called “<a href="https://infoseeking.org/iblog/2020/07/26/fate-report/">Google Or Not</a>.” This online game shows two sets of results for the same query. The objective is simple – pick the set that is reliable, trustworthy or most relevant.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A screenshot showing two sets of Google search results side-by-side" src="https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=463&fit=crop&dpr=1 600w, https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=463&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=463&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=582&fit=crop&dpr=1 754w, https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=582&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/376864/original/file-20201231-49525-gvvswc.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=582&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">In tests, about half the time people can’t tell the difference between Google search results containing misinformation and those with only trustworthy results.</span>
<span class="attribution"><span class="source">Chirag Shah</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>One of these two sets has one or two results that are either verified and labeled as misinformation or a debunked story. We made the game available publicly and advertised through various social media channels. Overall, we collected 2,100 responses from over 30 countries. </p>
<p>When we analyzed the results, we found that <a href="https://infoseeking.org/iblog/2020/10/23/people-cant-identify-covid-19-fake-news/">about half the time people mistakenly picked as trustworthy the set with one or two misinformation results</a>. Our experiments with hundreds of other users over many iterations have resulted in similar findings. In other words, about half the time people are picking results that contain conspiracy theories and fake news. As more people pick these inaccurate and misleading results, the search engines learn that that’s what people want.</p>
<p>Questions of Big Tech regulation and self-regulation aside, it’s important for people to understand how these systems work and how they make money. Otherwise market economies and people’s natural inclination to be attracted to eye-catching links will keep the vicious cycle going.</p>
<p>[<em>Understand new developments in science, health and technology, each week.</em> <a href="https://theconversation.com/us/newsletters/science-editors-picks-71/?utm_source=TCUS&utm_medium=inline-link&utm_campaign=newsletter-text&utm_content=science-understand">Subscribe to The Conversation’s science newsletter</a>.]</p><img src="https://counter.theconversation.com/content/152155/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Chirag Shah has received funding from National Science Foundation (NSF), National Institute of Health (NIH), Institute of Museum and Library Services (IMLS), Amazon, Google, and Yahoo! The view expressed here are his only.</span></em></p>Search engines, like social media algorithms, get you to click on links by learning what other people click on. Enticing misinformation often comes out on top.Chirag Shah, Associate Professor of Information Science, University of WashingtonLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1540602021-01-29T01:25:48Z2021-01-29T01:25:48ZIf Google does pull its search engine out of Australia, there are alternatives<figure><img src="https://images.theconversation.com/files/381048/original/file-20210128-19-1lmbpzf.jpg?ixlib=rb-1.1.0&rect=80%2C440%2C5029%2C3377&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock/Wachiwit</span></span></figcaption></figure><p>The Australian government’s <a href="https://www.theguardian.com/media/2020/dec/09/australia-is-making-google-and-facebook-pay-for-news-what-difference-will-the-code-make">push</a> to make Google pay news organisations for linking to their content has seen the search giant threaten to <a href="https://about.google/google-in-australia/an-open-letter/">pull out of Australia</a>. </p>
<p>Google Australia’s managing director Mel Silva said if the government’s proposal goes ahead, “we would have no real choice but to stop making Google Search available in Australia”.</p>
<p>Prime Minister Scott Morrison pushed back saying he won’t respond to “<a href="https://thenewdaily.com.au/news/politics/2021/01/22/pm-responds-google-threat/">threats</a>”. Even the Council of Small Business Organisations Australia says Google needs “<a href="https://www.theguardian.com/technology/2021/jan/27/ignore-threats-to-shut-search-in-australia-and-force-google-to-pay-small-business-groups-say">strong and stringent</a>” regulation because of its monopoly on searching the web.</p>
<h2>What if Google pulls out?</h2>
<p>Google’s proposal to make Google Search unavailable in Australia means we would need to search the web using other systems and tools. If this really happens, we could no longer go to <a href="https://www.google.com/">google.com</a> and <a href="https://www.google.com.au/">google.com.au</a> to search the web.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/its-not-fair-and-it-wont-work-an-argument-against-the-accc-forcing-google-and-facebook-to-pay-for-news-145391">It's not 'fair' and it won't work: an argument against the ACCC forcing Google and Facebook to pay for news</a>
</strong>
</em>
</p>
<hr>
<p>It is important to note that Google is not just web search. Google’s parent company <a href="https://abc.xyz/">Alphabet Inc</a> also runs key web portals such as YouTube, and productivity tools such as Gmail, Google Calendar, Google Docs and Google Maps (which actually <a href="https://www.businessinsider.com.au/google-maps-australia-tech-inventions-2020-2">started in Australia</a>). Those services are not going to be removed from the Australian market, even if web search does get pulled out.</p>
<p>Online advertising is another sector in which Google is the market leader and where it makes money. Pulling Google web search out from Australia does not mean businesses would no longer be able to advertise using Google’s services. </p>
<p>But with no Google Search here, those adverts would no longer appear ahead of any other search results and be visited by Australian users.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A Google Search result showing an ad for The Conversation ahead of any search results." src="https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=201&fit=crop&dpr=1 600w, https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=201&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=201&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=253&fit=crop&dpr=1 754w, https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=253&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/381044/original/file-20210128-23-guuupy.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=253&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Google Search places paid advertising ahead of any search results.</span>
<span class="attribution"><span class="source">Google.com/screenshot</span></span>
</figcaption>
</figure>
<p>Businesses would still be able to put their adverts on other Australian websites that use the <a href="https://ads.google.com/">Google Ads</a> service.</p>
<p>The issue with this scenario is that Google’s key competitive advantage is the ability to access <a href="https://www.wired.com/story/google-tracks-you-privacy/">data from people</a> using its search services. Pulling web search out from the Australian market would mean Google missing out on that data from people in Australia.</p>
<h2>The alternatives to Google</h2>
<p>Google is the dominant search engine in Australia — it has <a href="https://gs.statcounter.com/search-engine-market-share/all/australia">94% of the web search market in Australia</a> — but there are other search services.</p>
<p>The <a href="https://gs.statcounter.com/search-engine-market-share/all/australia">second most popular</a> search engine in Australia is <a href="https://www.bing.com/">Bing</a>, developed by Microsoft and often integrated into other Microsoft products such as its Windows operating system and Office tools.</p>
<p>Another less popular search option is <a href="https://au.yahoo.com/">Yahoo</a>, which also offers its own news and email service.</p>
<p>Other alternatives include niche search engines that offer unique tools with special features.</p>
<p>For example, <a href="https://duckduckgo.com/">DuckDuckGo</a> is a search engine that has recently risen in popularity thanks to a commitment to protecting its users’ privacy.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="The DuckDuckGo homepage" src="https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=303&fit=crop&dpr=1 600w, https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=303&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=303&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=381&fit=crop&dpr=1 754w, https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=381&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/381224/original/file-20210128-15-8jrbh.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=381&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">DuckDuckGo is gaining support.</span>
<span class="attribution"><a class="source" href="https://duckduckgo.com/">DuckDuckGo/Screen shot</a></span>
</figcaption>
</figure>
<p>Contrary to the web search products from Google and Microsoft, DuckDuckGo does not store its users’ search queries or track their interactions with the system.</p>
<p>The quality of DuckDuckGo’s search results has improved over time, and is now <a href="https://www.theguardian.com/technology/askjack/2019/dec/12/duckduckgo-google-search-engine-privacy">comparable</a> to that of the most popular search engines.</p>
<p>It says it now processes a daily average of more than <a href="https://duckduckgo.com/traffic">90 million search queries</a>, up from just over 51 million the same time last year.</p>
<p>Despite not drawing on users’ data to refine its search algorithms, the technology behind DuckDuckGo and other smaller players is based on the same machine-learning methods that others are using.</p>
<h2>Search the web, save the planet</h2>
<p>Another interesting and recent proposal of an alternative web search engine is <a href="https://www.ecosia.org/">Ecosia</a>. This system is unique as it focuses on sustainability and positive climate impact.</p>
<p>Its mission is to reinvest the income generated by search advertisements (the same business model Google Search is using) to <a href="https://info.ecosia.org/">plant trees</a> in key areas around the world.</p>
<p>So far, it says it has 15 million users and has contributed to planting more than 100 million trees, about 1.3 every second.</p>
<h2>Will Google really abandon Australia?</h2>
<p>Tim Berners-Lee, widely regarded as the inventor of the web, has <a href="https://www.theguardian.com/media/2021/jan/20/australias-proposed-media-code-could-break-the-world-wide-web-says-the-man-who-invented-it">pointed out</a> that the idea of asking web platforms to pay to post links runs counter to his fundamental concept.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/webs-inventor-says-news-media-bargaining-code-could-break-the-internet-hes-right-but-theres-a-fix-153630">Web's inventor says news media bargaining code could break the internet. He's right — but there's a fix</a>
</strong>
</em>
</p>
<hr>
<p>That said, it is also unfair for a search engine to make money using content that others have created.</p>
<p>It is also true that most of Google’s revenue already comes from asking others to pay for links on the web. This is <a href="https://ads.google.com/intl/en_au/home/how-it-works/">how Google’s online advertising works</a>: Google Ads makes advertisers pay for every impression users get or click users make to navigate to the advertised web page.</p>
<p>In some cases, if users end up <a href="https://support.google.com/google-ads/answer/1722022?hl=en-AU&ref_topic=3119146">buying the advertised product</a>, Google gets a payment.</p>
<p>More likely than Google pulling out of the Australian market, the government and the search giant should diplomatically find a compromise in which Google still provides its web search product in Australia and there will be a return to news organisations for Google making use of their content.</p>
<hr>
<p><em>This article was updated to clarify how Google can receive payments from advertisers for purchases.</em></p><img src="https://counter.theconversation.com/content/154060/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Gianluca Demartini receives funding from the Australian Research Council and Facebook.</span></em></p>There are other ways to search the web without Google, and some options help protect your privacy while another is good for the planet.Gianluca Demartini, Associate professor, The University of QueenslandLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1511672020-12-02T13:01:22Z2020-12-02T13:01:22ZCan an Apple search engine ever compete with Google?<figure><img src="https://images.theconversation.com/files/372042/original/file-20201130-13-1m5jpp4.jpg?ixlib=rb-1.1.0&rect=23%2C0%2C5152%2C2956&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">"Apple it" just doesn't have the same ring...</span> <span class="attribution"><a class="source" href="https://unsplash.com/photos/O5v8heKY4cI">Nathana Rebouças/unsplash</a>, <a class="license" href="http://artlibre.org/licence/lal/en">FAL</a></span></figcaption></figure><p>Small corners of the internet are ablaze with the news that Apple has significantly ramped up its <a href="https://www.macrumors.com/guide/applebot/">search bot activity</a>. Search bots typically scan websites in order to rank and index them for search engine results. When you look for something on a search engine, the results that appear are ordered by “ranking”, meaning that the result that is most accurate to what you are looking for appears at the top.</p>
<p>This increase in activity also appears alongside pressure from the UK competition commission to break up Apple’s multi-billion dollar <a href="https://www.theverge.com/2020/7/1/21310591/apple-google-search-engine-safari-iphone-deal-billions-regulation-antitrust">sweetheart deal</a> with Google. The deal ensures that Google is the default search engine for Apple’s iOS devices. <a href="https://www.ft.com/content/fd311801-e863-41fe-82cf-3d98c4c47e26">Many</a> are now anticipating that Apple is on track to launch its own search engine soon. </p>
<p>Apple’s entry in to the search engine market comes 11 years after Microsoft’s Bing made its debut, the only other noteworthy competitor to Google to date. Bing is by no means a success story, despite what Microsoft’s <a href="https://www.thurrott.com/cloud/microsoft-consumer-services/bing/133056/microsoft-bing-bigger-think">PR team</a> claims, and it continues to pale in comparison in terms of economic performance and market power to Google’s ubiquitous search platform. In fact, every academic year I and my colleague, Dr Kamal Munir, teach a case on Bing to the Cambridge MBA as a cautionary tale of what happens if you choose to go head on against entrenched platforms. Unlike Microsoft’s Bing, however, Apple’s opening gambit is quite different and is likely to produce a better outcome.</p>
<h2>Resetting the rules of engagement</h2>
<p>One of the bigger mistakes Microsoft made with the launch of Bing was to follow the same ad-based business model that Google was using. In this business model, search users enter what they are looking for and based on that the search engine also shows relevant ads that might interest them. For such a business to be profitable, you need a very large number of users searching, as well as a large number of advertisers willing to sell to them, alongside millions of websites scanned by the previously mentioned search bots. All three are needed to display useful search results for the user and bring the right customer to the advertiser. </p>
<p>In between, Google gets paid for matching the right ads with the right users. The more searches are made, the more useful the results. The more useful the results, the better is the ad targeting. Bing struggled to get this virtuous cycle started, and never really got to the scale that Google enjoys with its search offering.</p>
<p>Apple’s search engine will have a different future if rumours about its business model are true. Apple has been focusing heavily on <a href="https://www.apple.com/uk/privacy/features/">user privacy</a> recently, including but not limited to, publicly refusing to give secret access to its devices to the <a href="https://www.wired.com/story/the-time-tim-cook-stood-his-ground-against-fbi/">FBI</a>. It will be very much in line with this “privacy first” position that Apple chooses not to make money from advertising, which involves exposing customer usage data to third parties. Instead, it could simply sell more of its highly profitable devices and subscriptions to privacy-conscious customers. By not following Google’s footsteps, Apple does not have to engage with the search giant on its terms. </p>
<h2>The better product fallacy</h2>
<figure class="align-center ">
<img alt="An apple computer, laptop, watch and phone sit on their boxes in a neat order" src="https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=750&fit=crop&dpr=1 600w, https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=750&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=750&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=943&fit=crop&dpr=1 754w, https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=943&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/372541/original/file-20201202-22-ududgm.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=943&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Apple already has a plethora of popular products to supplement their earnings.</span>
<span class="attribution"><a class="source" href="https://unsplash.com/photos/Bs-zngH79Ds">Julian O'hayon/unsplash</a>, <a class="license" href="http://artlibre.org/licence/lal/en">FAL</a></span>
</figcaption>
</figure>
<p>When Bing was launched, it had <a href="https://news.microsoft.com/announcement/bing-search-engine-launches/">features</a> that Google did not have at that time. This included a “hover preview” of search results as well as a specialisation of sorts in travel, shopping, local businesses, and health searches. In terms of quality of search results, Microsoft’s <a href="https://techcrunch.com/2012/09/06/bing-it-on-microsoft-claims-21-users-prefer-bings-search-results-over-googles-in-blind-test/">claimed</a> similar or better output to Google’s. Despite its apparent product superiority, Bing never won the search engine wars.</p>
<p>This aspect of search engine history plays in favour of Apple, which does not need to differentiate itself from Google. In fact, Apple’s search results need to be “just good enough” to be adopted by its users en masse. We can see this with the results of Apple maps, which was launched back in 2012. Despite a publicly rocky launch thanks to its poor geographic coverage, Apple maps gained a <a href="https://www.theguardian.com/technology/2013/nov/26/apple-maps-europe-google">dominant market share</a> of 60% in the UK’s iPhone users in just under a year of launch. The same holds true for Apple music, which has become the <a href="https://www.macrumors.com/2020/04/03/apple-music-second-largest-streaming-service-2019/">second-biggest player</a> in streaming music despite a nine-year lead by Spotify.</p>
<h2>Disruptive side effects</h2>
<p>With its latest iOS 14 update, Apple has already started swapping out Google <a href="https://www.dailymail.co.uk/sciencetech/article-8888529/Apple-starts-displaying-search-results-iOS-14-latest-build-Google-killer.html">search results</a> in favour of its own. Most iOS users have barely noticed the change for all the reasons given above. But this silent swapping does not come without its own set of challenges. By defaulting to its search engine instead of Google on its devices, Apple will open itself to monopoly criticism from competition commissions in a variety of markets. It is also likely to upset the advertising industry who could lose their reach to Apple customers. The Apple customer base is a coveted one thanks to its better than average buying power, and by making it easier for users to avoid search ads, Apple might just create a <a href="https://www.emarketer.com/content/apple-upends-in-app-advertising-with-ios-14">tectonic shift</a> in the advertising industry as a whole.</p>
<p>Google’s dominance on internet search will not come to an end with Apple’s entry into the foray, but it would definitely weaken in the face of increasing consumer preference for privacy. Given that Google’s business model differs dramatically from Apple’s, it is likely that the search giant would have to learn to uncomfortably live with its rival’s search engine instead of pivoting to compete with it head on.</p><img src="https://counter.theconversation.com/content/151167/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Hamza Mudassir does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Will Apple’s search engine have the same ill fate as Microsoft’s Bing?Hamza Mudassir, Visiting Fellow in Strategy, Cambridge Judge Business SchoolLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1486912020-10-29T12:30:53Z2020-10-29T12:30:53ZGoogle antitrust case suggests Apple should be in the Department of Justice’s crosshairs too<figure><img src="https://images.theconversation.com/files/366276/original/file-20201028-13-1fpcvkw.jpg?ixlib=rb-1.1.0&rect=10%2C318%2C3327%2C2376&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Apple devices drive over half of all Google search traffic. </span> <span class="attribution"><a class="source" href="https://newsroom.ap.org/detail/AppleBeatsExxon/a953a0fa656b4464b5e0c8b8e18bea27/photo?Query=apple%20AND%20logo&mediaType=photo&sortBy=arrivaldatetime:desc&dateRange=Anytime&totalCount=270&currentItemNo=93">AP Photo/Russel A. Daniels</a></span></figcaption></figure><p>Google’s payments to Apple to promote its search engine in iPhones, iPads and Mac computers <a href="https://www.justice.gov/opa/pr/justice-department-sues-monopolist-google-violating-antitrust-laws">are at the center</a> of the Department of Justice’s antitrust lawsuit against the tech giant. </p>
<p>The suit alleges this creates a “continuous and self-reinforcing cycle of monopolization” by limiting which search engines consumers can use. </p>
<p>But as <a href="https://scholar.google.com/citations?user=z9oUtFsAAAAJ&hl=en&oi=ao">someone who studies platform markets, competition and industry structure</a>, I believe the agreement seems more like a damning indictment of Apple’s own potentially illegal business practices. </p>
<h2>Why Google needs Apple</h2>
<p>The Department of Justice alleges that Google pays Apple and other device-makers to set its search engine as the default “on billions of mobile devices and computers worldwide,” thus controlling how users access the internet. </p>
<p>It’s true, Google is dominant in search, which accounted for an <a href="https://www.forbes.com/sites/greatspeculations/2019/12/24/is-google-advertising-revenue-70-80-or-90-of-alphabets-total-revenue/#18092d894a01">estimated 83% of parent company Alphabet’s revenue</a> in 2019. </p>
<p>But <a href="https://www.businessinsider.com/apple-search-deal-google-value-privacy-2020-10">about half of Google’s search traffic</a> originates from Apple devices. If Apple were to replace Google with an alternative default search engine on its devices, I estimate that Google could lose US$30 billion to $40 billion in annual revenue, assuming most users didn’t change the setting back to Google.</p>
<p>Even if Apple didn’t pick a default and pushed the search engine choice to users, it would still have to create a list of possibilities. Research on <a href="https://www.doi.org/10.1287/ijoc.2020.0968">search</a> and <a href="https://scholar.smu.edu/cgi/viewcontent.cgi?article=1565&context=jalc">airline tickets</a> has shown that consumers overwhelmingly tend to pick whatever is at the top of the list, meaning Apple would still wield significant power over user choice. </p>
<p>Because of this, Google clearly has a powerful motive to keep its search engine as the default choice. </p>
<figure class="align-center ">
<img alt="The Google search application is pictured running on an iPhone." src="https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&rect=132%2C111%2C4530%2C2992&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=399&fit=crop&dpr=1 600w, https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=399&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=399&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=502&fit=crop&dpr=1 754w, https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=502&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/366230/original/file-20201028-13-1dzjqne.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=502&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Ultimately Google depends on device-makers like Apple to reach users.</span>
<span class="attribution"><a class="source" href="https://www.gettyimages.com/detail/news-photo/the-google-search-application-is-seen-running-on-an-iphone-news-photo/1027366886">Jaap Arriens/NurPhoto via Getty Images</a></span>
</figcaption>
</figure>
<h2>Why Apple would pick Google anyway</h2>
<p>Apple’s role as the gateway to billions of searches is the critical factor here. </p>
<p>Consider an Apple executive preparing the iPhone or another device for launch, choosing whether to set a default search engine and, if so, which one to pick. Presumably, there are two key factors: costs and customer satisfaction. </p>
<p>The cost to Apple of presetting a default search engine is negligible, just a few lines of code. Without a default, consumers would need to set it themselves or type google.com or bing.com themselves to conduct a search, as opposed to the common practice of typing a search term in the URL field. </p>
<p>To prevent this <a href="http://ceur-ws.org/Vol-2220/06_CONFWS18_paper_1.pdf">user inconvenience</a>, Apple would be best off presetting a search engine that was, ideally, the preferred choice of most users. The question then is: What would they prefer?</p>
<p>Google became synonymous with search since its founding in 1998 not simply due to its dominance – and <a href="https://www.nytimes.com/2011/10/16/technology/default-choices-are-hard-to-resist-online-or-not.html">payments to browser companies over the years</a> – but because users <a href="https://www.forbes.com/sites/forbesagencycouncil/2017/06/05/how-google-came-to-dominate-search-and-what-the-future-holds/#a3a0dbd38721">found the results of its algorithm and simple interface superior</a> to the competition. And Google <a href="https://www.statista.com/statistics/185966/us-customer-satisfaction-with-google/">continues to score high marks</a> with consumers in satisfaction surveys. </p>
<p>If Apple product managers were to preset one default search engine in order to maximize user satisfaction, they would probably pick Google anyway. </p>
<h2>A credible threat</h2>
<p>So why would Google pay <a href="https://www.nytimes.com/2020/10/25/technology/apple-google-search-antitrust.html">Apple $8 billion to $12 billion a year</a>? </p>
<p>In my view, it comes down to the the fear of being supplanted by a rival search engine if it stopped paying the fee. Apple has done this to Google before. </p>
<p>The iPhone used to come preloaded with two Google apps: Maps and <a href="https://www.scientificamerican.com/article/who-killed-youtube-on-the-iphone-apple-or-google/">YouTube</a>. In 2012, Apple kicked both off its devices as the two companies <a href="https://www.theguardian.com/technology/2012/nov/05/apple-google-maps-iphone-dropped">began to compete more aggressively with one another</a>, requiring consumers to download the apps if they wanted to use them. </p>
<p>From a game theory perspective, a <a href="https://www.econlib.org/library/Enc/GameTheory.html">credible threat or perception of one</a> could be enough to ensure continued compliance. </p>
<p><a href="https://www.emarketer.com/Article/Apple-Takes-Biggest-Bite-of-Mobile-Web-Traffic/1011234">Since at least 2014</a> – around <a href="https://www.nytimes.com/2020/10/25/technology/apple-google-search-antitrust.html">when the first Apple-Google partnership</a> on preset default occurred – Apple <a href="https://www.statista.com/statistics/236550/percentage-of-us-population-that-own-a-iphone-smartphone/">has dominated mobile web traffic</a>. This power gives Apple, as a platform providing access to users, the leverage it needs to charge and potentially extort a rent – in <a href="https://www.investopedia.com/terms/e/economicrent.asp">economic parlance</a> – for a product design decision that it would have likely chosen on its own. This could violate antitrust law, though Apple would likely argue it’s merely monetizing a resource it built.</p>
<h2>It all comes down to the platform</h2>
<p><a href="https://books.google.com/books?hl=en&lr=&id=Bvd1CQAAQBAJ">Platforms provide</a> the technological and economic infrastructure and set the rules participants must abide by. </p>
<p>This gives them significant power as the access point to potentially massive numbers of users, which has been the core issue underlying past antitrust actions against major tech companies such as <a href="https://www.justice.gov/atr/complaint-us-v-microsoft-corp">Microsoft in the late 1990s</a>.</p>
<p>While the Department of Justice lawsuit <a href="https://www.axios.com/heres-what-the-us-antitrust-case-charges-google-with-cbaf5458-0fd9-4c3d-a85e-0c8637675ea5.html">does have a strong case against Google in other areas</a>, it seems like the part about the Google-Apple partnership should be more directed toward the company that actually controls the access to consumers.</p>
<p>And with <a href="https://nypost.com/2020/10/28/apple-is-reportedly-looking-to-develop-its-own-search-engine/">new reports that Apple is planning</a> to develop its own search engine, the government’s desired remedy in its lawsuit – the end of the partnership and the Google default – may happen anyway, making the case mostly moot. </p>
<p>[<em>You’re smart and curious about the world. So are The Conversation’s authors and editors.</em> <a href="https://theconversation.com/us/newsletters/weekly-highlights-61?utm_source=TCUS&utm_medium=inline-link&utm_campaign=newsletter-text&utm_content=weeklysmart">You can get our highlights each weekend</a>.]</p><img src="https://counter.theconversation.com/content/148691/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Hemant K. Bhargava's work has been supported and inspired by various tech firms, including Google (and other search firms such as Yahoo! and Overture) with a research excellence gift from the Google Cloud Platform in 2018.</span></em></p>Google pays Apple to make its search engine the default on its devices, but the iPhone maker actually has more market power in the relationship.Hemant K. Bhargava, Professor, Suran Chair in Technology Management; Director, Center for Analytics and Technology in Society, University of California, DavisLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/968382018-06-04T10:37:05Z2018-06-04T10:37:05ZWhen will Google defend democracy?<figure><img src="https://images.theconversation.com/files/220802/original/file-20180529-80623-ieg905.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">How does searching affect voting?</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-vector/illustration-long-shadow-glass-magnifier-hand-658091983">Blablo101/Shutterstock.com</a></span></figcaption></figure><p>As the 2018 midterm elections approach in the U.S., Google’s <a href="http://www.politico.com/magazine/story/2015/08/how-google-could-rig-the-2016-election-121548.html">power to influence undecided voters</a> remains overshadowed by Facebook’s <a href="https://theconversation.com/understanding-facebooks-data-crisis-5-essential-reads-94066">personal data crisis</a>.</p>
<p>Facebook has “<a href="https://www.nytimes.com/2018/04/13/technology/facebook-silicon-valley.html">taken it on the chin</a>” for its role in the 2016 presidential election, and organizations like the <a href="https://www.nytimes.com/2018/03/17/us/politics/cambridge-analytica-trump-campaign.html">political consulting firm</a> <a href="https://theconversation.com/how-cambridge-analyticas-facebook-targeting-model-really-worked-according-to-the-person-who-built-it-94078">Cambridge Analytica</a> and the <a href="https://www.nytimes.com/interactive/2017/01/06/us/politics/document-russia-hacking-report-intelligence-agencies.html">Russian troll farm</a> known as the <a href="https://www.wired.com/story/mueller-indictment-internet-research-agency/">Internet Research Agency</a> have dominated headlines. Yet, despite having a <a href="https://www.dissentmagazine.org/online_articles/googles-snoops-mining-our-data-for-profit-and-pleasure">troubling history</a> and collecting <a href="https://www.nytimes.com/2018/05/16/technology/personaltech/google-personal-data-facebook.html">more personal data</a> through <a href="https://www.usnews.com/opinion/articles/2013/05/10/15-ways-google-monitors-you">more products</a> than Facebook, Google has somehow managed to evade the public spotlight on this one. That may be changing.</p>
<p>The U.S. Senate Judiciary Committee recently sent Google a <a href="https://www.judiciary.senate.gov/imo/media/doc/2018-04-10%20CEG%20to%20Google%20-%20Data%20Privacy.pdf">letter asking a series of questions</a> about the company’s personal data protections. As one of the researchers who helped discover that search engines can <a href="https://doi.org/10.1073/pnas.1419828112">substantially influence users’ voting preferences</a>, I found the last question to be the most intriguing: “Are you aware of any foreign entities seeking to influence or interfere with U.S. elections through your platforms?” If Google’s response to this question exists, it has not been made public. </p>
<h2>Search engine influence</h2>
<p>Since 2013, I’ve been involved in the design and execution of a <a href="http://aibrt.org/index.php/internet-studies">long series of experiments</a> that have demonstrated how search engines can influence undecided voters’ candidate choices through nearly undetectable manipulations to search rankings. We labeled this powerful new form of influence the <a href="https://doi.org/10.1073/pnas.1419828112">search engine manipulation effect</a>.</p>
<p>The way this effect works is simple: Favoritism for a particular candidate in election-related search rankings leads to people preferring that candidate. For example, a search related to an upcoming election might return results favoring candidate A higher than results favoring candidate B. That’s called partisan <a href="https://doi.org/10.1145/2998181.2998321">ranking bias</a>. Since people tend to <a href="https://doi.org/10.1111/j.1083-6101.2007.00351.x">click on and trust</a> highly ranked results, more people will then trust and consume the information supporting candidate A. In turn, that consumption increases their preference for candidate A. </p>
<p>The most important aspect of this effect, however, is that most people can’t detect the partisan ranking bias – and it’s virtually impossible to defend yourself from <a href="https://aeon.co/essays/how-the-internet-flips-elections-and-alters-our-thoughts">influences you can’t perceive</a>. Fortunately, in three follow-up experiments, involving 3,600 participants, we demonstrated that alerting people to partisan ranking bias <a href="https://doi.org/10.1145/3134677">can help suppress the effect</a> – though only laws or regulations actually preventing partisan ranking could eliminate the effect entirely. </p>
<h2>Why focus on Google?</h2>
<p>Google handles <a href="https://www.statista.com/statistics/267161/market-share-of-search-engines-in-the-united-states/">more than 60 percent</a> of internet search activity in the U.S., and nearly <a href="https://www.statista.com/statistics/216573/worldwide-market-share-of-search-engines/">90 percent worldwide</a>. Every year, this translates to <a href="https://searchengineland.com/google-now-handles-2-999-trillion-searches-per-year-250247">trillions of queries</a> related to people’s private thoughts, concerns and questions.</p>
<p>With respect to news, search engines are a <a href="https://reutersinstitute.politics.ox.ac.uk/sites/default/files/Digital%20News%20Report%202017%20web_0.pdf">bigger</a> <a href="https://www.americanpressinstitute.org/publications/reports/survey-research/how-americans-get-news/">source</a> than social media. Although an often cited 2016 Pew study found that a majority, <a href="http://www.journalism.org/2016/05/26/news-use-across-social-media-platforms-2016/">62 percent, of U.S. adults</a> got news on social media, the devil is in the details. If you unpack that statistic, you’ll find that 18 percent do so “hardly ever.” Added to the 38 percent of Americans who “never” got news on social media, the same study suggests that social media is a negligible source of news for 56 percent of Americans, also a majority.</p>
<p><img src="https://cdn.theconversation.com/static_files/files/138/PewGif.gif?1527691715"></p>
<p>Think about it: When you need to fact-check something or learn more about a topic, what do you do? You <a href="https://www.merriam-webster.com/dictionary/google">Google</a> it. This fact is supported by a recent international survey that found that <a href="https://dx.doi.org/10.2139/ssrn.2944191">74 percent of participants</a> reported using search engines to fact-check information they found on social media. The same survey found that <a href="https://dx.doi.org/10.2139/ssrn.2944191">68 percent</a> reported that the information they found while searching was “important to influencing their decisions about voting.” </p>
<h2>What does Google think?</h2>
<p>Google’s executives rarely make public responses to critiques of its search system. But in 2015, my mentor at the time <a href="http://drrobertepstein.com">Robert Epstein</a> published an article in Politico – entitled “<a href="http://www.politico.com/magazine/story/2015/08/how-google-could-rig-the-2016-election-121548.html">How Google Could Rig the 2016 Election</a>” – and that did the trick. Google’s head of search at the time, Amit Singhal, responded with his own article, calling Epstein a <a href="https://www.politico.com/magazine/story/2015/08/google-2016-election-121766">conspiracy theorist</a>, stating that “there is absolutely no truth to Epstein’s hypothesis that Google could work secretly to influence election outcomes” and that “Google has never ever re-ranked search results on any topic (including elections) to manipulate user sentiment.”</p>
<p>Singhal’s first claim is hard to believe, unless you dismiss <a href="https://doi.org/10.1073/pnas.1419828112">our research</a>, <a href="https://doi.org/10.1145/3134677">our replication</a>, and the <a href="https://doi.org/10.1145/3121050.3121074">independent research</a> <a href="https://doi.org/10.1145/3020165.3020185">built on our findings</a>. Search engines do have the capacity to shift people’s opinions, including who to vote for.</p>
<p>His second claim, that Google “never ever re-ranked search results,” also doesn’t quite hold up: The EU recently fined the company <a href="https://www.nytimes.com/2017/06/27/technology/eu-google-fine.html">US$2.7 billion</a> for ranking its own services higher in search results than its competitors.</p>
<h2>Defending democracy</h2>
<p>Another one of the <a href="https://www.judiciary.senate.gov/imo/media/doc/2018-04-10%20CEG%20to%20Google%20-%20Data%20Privacy.pdf">Judiciary Committee’s questions to Google</a> also struck a chord with me: “How do you monitor the ability of foreign entities to influence and interfere with U.S. elections?” </p>
<p>This question struck me because I’ve been developing systems for exactly this purpose – preserving search rankings and analyzing them for systematic differences – for <a href="https://doi.org/10.1145/3178876.3186143">several</a> <a href="http://aibrt.org/downloads/EPSTEIN_&_ROBERTSON_2017-A_Method_for_Detecting_Bias_in_Search_Rankings-AIBRT_WP-17-02_6-1-17.pdf">years</a>. In the course of this work, however, I’ve come to believe that freeing the democratic process from technologically enabled influences is virtually impossible without the cooperation of modern tech giants.</p>
<p>Facebook is now <a href="https://newsroom.fb.com/news/2018/04/new-elections-initiative/">offering to collaborate with academic researchers</a> who can measure and perhaps lessen or prevent undue influence on elections, and <a href="http://www.niemanlab.org/2018/04/facebook-and-twitter-are-opening-up-a-bit-to-academic-researchers-so-platforms-can-make-better-decisions/">Twitter is doing something similar</a>. Related efforts are also bringing transparency to other platforms like <a href="https://www.nytimes.com/2018/03/10/opinion/sunday/youtube-politics-radical.html">YouTube</a> and <a href="https://www.reddit.com/r/announcements/comments/8bb85p/reddits_2017_transparency_report_and_suspect/">Reddit</a>. When will Google get on board?</p>
<p>At an <a href="https://apen18.hans-bredow-institut.de/program/">upcoming conference</a>, I will <a href="http://ronalderobertson.com/static/robertson_icwsm2018_final.pdf">present the latest system</a> I’ve been designing with <a href="https://cbw.sh">Christo Wilson</a>, a leading scientist in the field of algorithm auditing, for monitoring search rankings for partisan bias. With a little assistance from Google, no more than Facebook is offering, accurately monitoring or preventing search engine influence in the 2018 elections is actually a feasible goal. Without the company’s help, things look bleak.</p>
<p>Although Google is an advertising business, its core is composed of creative and intelligent individuals who care deeply about the impact their work has on the world. This is evidenced by <a href="https://www.nytimes.com/2018/04/04/technology/google-letter-ceo-pentagon-project.html">the recent letter</a> signed by more than 3,100 Google employees protesting the use of their work in warfare technology. Nearly a dozen Google workers went so far as to <a href="https://gizmodo.com/google-employees-resign-in-protest-against-pentagon-con-1825729300">resign in protest</a>.</p>
<p>Perhaps the day is fast approaching when Google will step up, as Facebook, Twitter and Reddit have, to help defend democracy from the new world of computational propaganda. Perhaps there is already a letter circulating internally and gathering signatures. With state and federal primary elections already underway, let’s hope so.</p>
<p><em>Correction: This article was updated on June 11, 2018, to correct the description of Robert Epstein.</em></p><img src="https://counter.theconversation.com/content/96838/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Ronald Robertson is supported in part by NSF grants IIS-1408345 and IIS-1553088. Any opinions, findings, and conclusions
or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the NSF.</span></em></p>Social media sites aren’t the only online systems that can secretly influence people’s votes. Search engines can too and may be even more successful – and undetectable.Ronald Robertson, Ph.D. Student in Network Science, Northeastern UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/791622017-08-03T01:00:12Z2017-08-03T01:00:12ZWhy the creators of ‘13 Reasons Why’ should pay attention to the spike in suicide-related Google searches<figure><img src="https://images.theconversation.com/files/180760/original/file-20170802-7625-bewz2t.png?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">As the show's popularity surged, interest in suicide also grew.</span> <span class="attribution"><span class="source">Nick Lehr/The Conversation via www.shutterstock.com</span>, <a class="license" href="http://creativecommons.org/licenses/by-nc-sa/4.0/">CC BY-NC-SA</a></span></figcaption></figure><p>Does it matter that people seem to have become more interested in suicide – expressing more suicidal thoughts, while becoming more likely to research ways to commit or prevent suicide – in the wake of the popular Netflix series “13 Reasons Why”?</p>
<p>According to <a href="http://jamanetwork.com/journals/jamainternalmedicine/fullarticle/10.1001/jamainternmed.2017.3333">new research</a> my colleagues and I conducted, suicide-related Google searches increased in the weeks following the spring release of the popular Netflix series “13 Reasons Why.” </p>
<p>The show – which became the streaming service’s <a href="http://www.teenvogue.com/story/13-reasons-why-netflix-most-popular-show-social-media">most discussed series on social media</a> – chronicles a high school girl’s suicide over the course of 13 episodes. In the season finale, the suicide is depicted in a three-minute scene.</p>
<p>Singer Selena Gomez, the executive producer of the show, <a href="http://www.rollingstone.com/culture/news/selena-gomez-defends-13-reasons-why-as-honest-w486466">said</a> she hoped the series would raise suicide awareness. Yet some – including <a href="https://www.schoolcounselor.org/school-counselors-members/professional-development/learn-more/13-reasons-why-resources">educators and school psychologists</a> – fear the series glamorizes the victim and her suicide in a way that could promote copycat behavior. </p>
<p>In recent months, stories about the possible effects of the series have circulated in the media. In May, The Washington Post <a href="https://www.washingtonpost.com/local/education/educators-and-school-psychologists-raise-alarms-about-13-reasons-why/2017/05/01/bb534ec6-2c2b-11e7-a616-d7c8a68c1a66_story.html?utm_term=.d44ed16c1294">reported</a> that school administrators in Florida were witnessing more risky behaviors among their students, from self-mutilation to suicide threats. In June, People magazine <a href="http://people.com/chica/peruvian-man-imitates-13-reasons-why-commits-suicide/">told the story</a> of a young man in Peru who took his own life, leaving behind recordings in a way that mimicked the main character in the series. </p>
<p>Our study adds to this discussion by considering a novel source: search data. Because the internet is a place where people can anonymously search for information free of judgment, researchers can see what’s on the mind of the public by monitoring what they’re searching. </p>
<p>For this reason, my colleagues and I will often turn to the internet to track real-time trends in <a href="https://well.blogs.nytimes.com/2013/04/11/a-seasonal-pattern-to-mental-health/">mental heath</a> and <a href="https://well.blogs.nytimes.com/2014/01/08/hard-times-and-headaches/">other health concerns</a> in order to better understand how the public is thinking, feeling and behaving. (For example, earlier this year we demonstrated that Charlie Sheen’s HIV disclosure in 2015 corresponded with record levels of <a href="https://tonic.vice.com/en_us/article/wnwkab/the-charlie-sheen-effect-on-hiv-testing-was-equivalent-to-7-world-aids-days">interest in HIV testing</a>, which showed search trends could signal concrete prevention outcomes.)</p>
<p>With “13 Reasons Why,” we wanted to see how the content and volume of internet searches about suicide changed after the series’ release. We analyzed suicide-related searches from March 31, the day the series was released, to April 18, the day before former NFL star Aaron Hernandez <a href="https://www.bostonglobe.com/metro/2017/04/19/aaron-hernandez-kills-himself-prison/Hkp9wdGcZImoMBomJLMNVJ/story.html">committed suicide in prison</a> (a national news event that likely caused suicide-related searches to spike on its own). Then we compared these results to the expected search volumes had the series never been released, figures we arrived at by analyzing daily search trends between January 15 and March 30. </p>
<p>Compared to the searches between January and March, searches that signaled suicide awareness (terms such as “suicide prevention”) and suicidal ideation (terms such as “how to commit suicide”) increased following the release of “13 Reasons Why.” People were also more likely to use search terms like “teen suicide,” “suicidal thoughts” and “how to kill yourself.” </p>
<p>Together, suicide searches were 19 percent higher for the 19 days following the series’ release compared to the period before the release. That figure reflects over one million more searches than what would have normally been expected. </p>
<p>While heightened suicide awareness can be a good thing, the spike in searches that indicate suicidal ideation, from ways to commit suicide to suicidal thoughts, could signal a more disturbing trend. <a href="http://www.jad-journal.com/article/S0165-0327(11)00052-8/fulltext">Prior research</a> has shown that suicide search trends are correlated with actual suicides, and <a href="http://jech.bmj.com/content/57/4/238">media coverage</a> of suicides concur with increased suicide attempts.</p>
<p>While we don’t know whether any specific search preceded an actual suicide attempt, the trends we found in search data suggest that the show’s creators probably have an obligation to mitigate suicidal ideation in the future. </p>
<p>Netflix did give “13 Reasons Why” a TV-MA rating, and a few episodes did have specific warnings for explicit material.</p>
<p>But if they want to go further, they could follow the <a href="http://www.who.int/mental_health/prevention/suicide/resource_media.pdf">World Health Organization’s media guidelines</a> for preventing suicide by removing the scenes that depict the actual suicide. They could also display suicide hotline numbers at the start of each episode. These suggestions could be retrofitted to season one and considered prior to the release of the <a href="http://www.eonline.com/news/860730/13-reasons-why-season-2-officially-underway-get-your-first-look">second season</a>. </p>
<p>Either way, the findings underscore the value of big data from online social systems. Whether it’s searches about suicide, HIV testing or other health concerns, the rapid and reliable information from search engines can make public health agencies more responsive to the populations they serve.</p><img src="https://counter.theconversation.com/content/79162/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Jon-Patrick Allem does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Researchers found that suicide-related searches soared in the weeks after the show’s release. What are the broader public health implications?Jon-Patrick Allem, Research Scientist, University of Southern CaliforniaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/785702017-07-20T01:50:11Z2017-07-20T01:50:11ZThe Library of Congress opened its catalogs to the world. Here’s why it matters<figure><img src="https://images.theconversation.com/files/176956/original/file-20170705-28939-r7aodn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">The Library of Congress is in Washington, D.C.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/washington-dc-usa-august-5-2016-570867745?src=lGpFCvv5qT3HOH2r3LGHFA-1-44">Valerii Iavtushenko/Shutterstock.com</a></span></figcaption></figure><p>Imagine you wanted to find books or journal articles on a particular subject. Or find manuscripts by a particular author. Or locate serials, music or maps. You would use a library catalog that includes facts – like title, author, publication date, subject headings and genre. </p>
<p>That information and more is stored in the treasure trove of library catalogs. </p>
<p>It is hard to overstate how important this library catalog information is, particularly as the amount of information expands every day. With this information, scholars and librarians are able to find things in a predictable way. That’s because of the descriptive facts presented in a systematic way in catalog records.</p>
<p>But what if you could also experiment with the data in those records to explore other kinds of research questions – like trends in subject matter, semantics in titles or patterns in the geographic source of works on a given topic?</p>
<p>Now it is possible. The Library of Congress has made <a href="https://web.archive.org/web/20170525222956/https://www.loc.gov/item/prn-17-068/">25 million digital catalog records</a> available for anyone to use at no charge. The free data set includes records from 1968 to 2014.</p>
<p>This is the largest release of digital catalog records in history. These records are part of a data ecosystem that crosses decades and parallels the evolution of information technology. </p>
<p>In my <a href="http://quod.lib.umich.edu/c/crmstoolkit">research about copyright and library collections</a>, I rely on these kinds of records for information that can help determine the copyright status of works. The data in these records already are embodied in library catalogs. What’s new is the free accessibility of this organized data set for new kinds of inquiry.</p>
<p>The decision reflects a fresh attitude toward shared data by the Library of Congress. It is a symbolic and practical manifestation of the library’s leadership aligned with its mission of public service.</p>
<h2>Some history</h2>
<p>To understand the implications of this news, it helps to know a bit about the <a href="https://americanlibrariesmagazine.org/2016/01/04/cataloging-evolves/">history of library catalog records</a>. </p>
<p>Today, search engines let us easily find books we want to borrow from libraries or purchase from any number of sources. Not long ago, this would have seemed magical. Search engines use data about books – like the title, author, publisher, publication date and subject matter – to identify particular books. That descriptive information was gathered over the years in library catalog records by librarians. </p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=502&fit=crop&dpr=1 754w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=502&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=502&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Card catalog at the Library of Congress.</span>
<span class="attribution"><a class="source" href="https://www.flickr.com/photos/rrenomeron/8493938344/">Rich Renomeron/flickr</a>, <a class="license" href="http://creativecommons.org/licenses/by-nc-nd/4.0/">CC BY-NC-ND</a></span>
</figcaption>
</figure>
<p>The library’s action sheds light on this unseen but critical network. This infrastructure is invisible to most of us as we use libraries, buy books or use search engines. </p>
<p>For many, the idea of a library catalog conjures up the image of card catalogs. The descriptions contained in catalog records are “metadata” – information about information. Early catalog records date back to 1791, just after the French Revolution. The revolutionary government used playing cards to document property seized from the church. The idea was to make a <a href="https://archive.is/20121215093454/http://gslis.simmons.edu/wikis/LIS415OL_History_Encyclopedia/Origins_of_the_Card_Catalog">national bibliography of library holdings</a> confiscated during the Revolution. </p>
<p>For many years, library collections were organized individually. As the number of books and libraries grew, the increased complexity demanded a more consistent approach. For example, when the <a href="https://www.loc.gov/loc/lcib/0806/jefferson.html">Library of Congress purchased Thomas Jefferson’s personal library</a> in 1815, it arranged its collections around Jefferson’s personal system organized around the themes of memory, reason and imagination. (Jefferson based this on <a href="http://www.loc.gov/exhibits/thomas-jeffersons-library/">Francis Bacon’s own model</a>.) The library sought to arrange its collections on that model into the 19th century. </p>
<figure class="align-left ">
<img alt="" src="https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Books on my shelf, marked with KF and HB. The K indicates that the book relates to law, the H that it relates to social science. The second letter indicates a subcategory.</span>
<span class="attribution"><span class="source">Melissa Levine</span>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span>
</figcaption>
</figure>
<p>As the number of books and libraries grew, a more systematic approach was needed. The Dewey Decimal System appeared in 1876 to tackle this challenge. It combined consistent numbers (“classes”) with particular topics. Each class can be further divided for more specific descriptions. </p>
<p>In the 1890s, the library developed the <a href="https://www.loc.gov/catdir/cpso/lcc.html">Library of Congress Classification System</a>. It is still used today to predictably manage millions of items in libraries worldwide.</p>
<h2>Catalogs, cards and computers</h2>
<p>By the 1960s, systematic descriptions made the transition from analog cards to online catalog systems a natural step. <a href="https://www.loc.gov/marc/umb/um01to06.html">Machine-Readable-Cataloging (or MARC) records</a> were developed to electronically read and interpret the data in bibliographic cataloging records. The structured categorization coincided naturally with the use of computers. </p>
<p>Now, <a href="http://lj.libraryjournal.com/2002/10/ljarchives/marc-must-die/#_">MARC records</a> too are on the way out, making room for <a href="https://www.loc.gov/bibframe/faqs/">more modern and flexible standards</a>. </p>
<p>The Library of Congress remains a primary – but not the only – source for catalog records. Individual libraries produce catalog records that are compiled and circulated through organizations like <a href="https://www.oclc.org/en/home.html">OCLC</a>. OCLC connects libraries around the globe and offers an online catalog. <a href="https://www.oclc.org/en/worldcat.html">WorldCat</a> coordinates catalog records from many libraries into a cohesive online resource. Groups like these charge libraries through membership fees for access to the compiled data. Libraries, though, typically do not charge for the catalog records they produce, instead working cooperatively through organizations like OCLC. This may evolve as more <a href="https://everybodyslibraries.com/2016/09/12/the-value-of-catalogs-in-the-linked-data-era-two-recent-talks/">shared effort and crowdsourced resources can be combined </a>with the library’s data in ways that improve search and inquiry. Examples include <a href="http://www.share-research.org/">SHARE</a> and <a href="https://www.wikipedia.org/">Wikipedia</a>. </p>
<h2>One month later</h2>
<p>In the short time since the Library of Congress’ data release, we see inklings of what may come. At a <a href="http://blogs.loc.gov/thesignal/2017/06/hack-to-learn-at-the-library-of-congress/?loclr=eadpb">Hack-to-Learn event</a> in May, researchers showed off early experiments with the data, including a <a href="https://medium.com/@thisismattmiller/library-of-congress-lists-57ddd177f1e2?loclr=blogsig">zoomable list of nine million unique titles</a> and <a href="http://www.vam.ac.uk/blog/digital-media/visualising-library-catalogues">a natural language interface with the data</a>. </p>
<p>For my part, I am considering how to use the library’s data to learn more about the history of publishing. For example, it might be possible to see if there are trends in dates of publication, locations of publishers and patterns in subject matter. It would be fruitful to correlate copyright information data retained by the U.S. Copyright Office to see if one could associate particular works with their copyright information like registration, renewal and ownership changes. However, those records remain in formats that remain difficult to search or manipulate. The <a href="https://www.copyright.gov/digitization/records.html">records prior to 1978</a> are not yet available online at all from the U.S. Copyright Office. </p>
<p><a href="https://www.lib.umich.edu/users/joque">Colleagues</a> at the University of Michigan Library are studying the recently released records as a way to practice map-making and explore geographic patterns with visualizations based on the data. They are thinking about gleaning locations from subject metadata and then mapping how those locations shift through time. </p>
<p>There’s a growing expectation that this kind of data should be freely available. This is evidenced by the expanding number of open data initiatives, from institutional repositories such as <a href="https://deepblue.lib.umich.edu/data/">Deep Blue Data</a> here at the University of Michigan Library to the U.S. government’s <a href="https://www.data.gov/open-gov/">data.gov</a>. The U.K.’s <a href="http://www.universitiesuk.ac.uk/policy-and-analysis/research-policy/open-science/Pages/open-research-data-task-force.aspx">Open Research Data Task Force</a> just released a <a href="http://www.universitiesuk.ac.uk/policy-and-analysis/research-policy/open-science/Documents/ORDTF%20report%20nr%201%20final%2030%2006%202017.pdf">report</a> discussing technical, infrastructure, policy and cultural matters to be addressed to support open data.</p>
<p>The Library of Congress’ action demonstrates an overarching shift in use of technology to meet historical research missions and advance beyond. Because the data are freely available, anyone can experiment with them.</p><img src="https://counter.theconversation.com/content/78570/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Melissa Levine has received funding from the Institute for Museum and Library Services.</span></em></p>Catalog data are a library’s most important map to knowledge. What does it mean that
the Library of Congress just released 25 million records to the public?Melissa Levine, Lead Copyright Officer, Librarian, University of MichiganLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/801292017-07-12T23:25:08Z2017-07-12T23:25:08ZOnline shopping: Retailers seek visibility in face of Google control<figure><img src="https://images.theconversation.com/files/177729/original/file-20170711-14431-1pinore.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Google argues that it simply gives consumers what they want.</span> <span class="attribution"><span class="source">(Shutterstock)</span></span></figcaption></figure><p>Customers often find retailers online using Google. For example, type “laptop” into the wildly popular search engine, and you will quickly see web links related to laptop computers. Some of those are “sponsored links,” also known as retailer ads. Those retailers paid Google to display their links in searches for that keyword.</p>
<p>This sponsored search advertising is popular with retailers and provides much of Google’s revenue. The tech behemoth took in some <a href="http://searchengineland.com/google-search-ad-revenues-271188">US$24 billion</a> in 2016 from the United States alone — about 76 per cent of the country’s search ad market.</p>
<p>That popularity means it’s important for online retailers to understand the advertising process. What factors help links appear first on the page? Are some retailers better at this critical competition for visibility?</p>
<p>Furthermore, some ads are for Google’s own retail site. Does that matter? Should we be concerned that Google has several ways to influence which ads we see?</p>
<h2>Competing for visibility</h2>
<p>Search advertising requires many decisions. Should retailers sponsor just a few keywords, or many? Which ones should they choose? How much money should they offer to pay Google for each word?</p>
<p>These decisions matter because customers click more on links near the top of the page. The first link displayed can have <a href="http://www.smartinsights.com/search-engine-marketing/search-engine-statistics/">double</a> the “click-through” rate of the second one.</p>
<p><a href="http://onlinelibrary.wiley.com/doi/10.1111/deci.12038/abstract">Research</a> at our school, The Goodman School of Business at Brock University, found that some firms are much better than others at managing this visibility challenge. Among online-only retailers like Amazon, the greatest differences are in search rankings. The best firms get their ads nearer the top at relatively lower costs.</p>
<p>Among retailers like Staples that are multi-channel (meaning they also have bricks-and-mortar stores), the biggest differences instead are in the rates that consumers click on links and buy products. The best firms get more clicks and more sales per ad dollar. Surprisingly, multi-channel retailers tend to be more efficient overall than the online-only ones at search advertising.</p>
<p>A second <a href="http://www.sciencedirect.com/science/article/pii/S0167923614002760">study</a> examined how popularity, payments and competition affect rankings. Not surprisingly, popular retailers tend to rank highly in search results. Retailers offering to pay Google more per word also rank higher.</p>
<h2>Payments to Google sway rankings</h2>
<p>Those payments have more influence on search rankings for obscure web pages than for popular ones. Popular sites tend to rank highly regardless of the payment they offer Google. </p>
<p>When multi-channel retailers like Staples face few competitors for keywords, popularity and payment have a large impact on their rankings. But when there’s a lot of competitors, multi-channel retailers tend to fall down the page. They seem to rely more then on their physical stores and less on the web.</p>
<p>Online-only retailers don’t have that option, and seem to sponsor links regardless of the competition. Popular online-only retailers generally appear high on the page no matter the number of competitors. Similarly, competition seemingly does not affect the payments they offer.</p>
<p>(It will be interesting to see how this distinction evolves as some online-only retailers add brick-and-mortar stores, as with Amazon’s <a href="https://theconversation.com/amazon-dives-into-groceries-with-whole-foods-five-questions-answered-79638">takeover</a> of Whole Foods. Or as traditional retailers move increasingly online, as with Walmart’s <a href="https://www.reuters.com/article/us-bonobos-m-a-walmart-idUSKBN1971RV">purchase</a> of online men’s retailer Bonobos.)</p>
<h2>Google’s wide reach</h2>
<p>Google also places ads elsewhere in its AdSense network, such as on the web pages you visit. It chooses those ads partly by tracking your previous searches. For example, if you searched for “laptops” earlier, you’ll likely see ads for laptops on subsequent websites.</p>
<p>Google also places ads in its free version of Gmail. It chooses those ads in part by scanning your emails’ contents. If you’ve been exchanging messages about laptops lately, you’ll likely see ads for them there. Google recently announced <a href="http://www.cbc.ca/news/business/google-gmail-ads-1.4178052">plans to end</a> that scanning, however, due to the confusion it’s caused for its paid Gmail users.</p>
<p>The company influences online ads’ visibility in other ways too. Its search engine <a href="https://theconversation.com/is-your-website-mobile-friendly-google-update-will-hit-you-badly-if-not-40571">prefers</a> sites that work well on <a href="https://theconversation.com/cyber-monday-gives-a-big-boost-to-mobile-commerce-68902">mobile</a> devices like smart phones. Its Chrome web browser will soon <a href="http://www.cbc.ca/news/technology/auto-play-video-ad-blockers-1.4158663">block</a> annoying ads that pop-up on the screen or that auto-play noisy videos.</p>
<figure class="align-right ">
<img alt="" src="https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=381&fit=crop&dpr=1 600w, https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=381&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=381&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=479&fit=crop&dpr=1 754w, https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=479&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/177726/original/file-20170711-14421-3vjxir.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=479&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Margrethe Vestager, the European Union’s competition commissioner, speaks at a news conference in Brussels in June. The EU has fined Google over its online shopping service.</span>
<span class="attribution"><span class="source">(AP Photo/Virginia Mayo)</span></span>
</figcaption>
</figure>
<p>Google has a lot of control over what we see online. It’s one of the <a href="https://plato.stanford.edu/entries/ethics-search/">ethical questions</a> that all search engines raise. An added concern here is that Google sells products through its own Shopping site. That means it competes with companies that it advertises, all while controlling whose ads are easiest to see.</p>
<p>When European Union regulators investigated this potential conflict of interest, they found that Google Shopping links tend to appear much higher in Google searches than those of rival marketplaces. The EU is therefore <a href="http://www.cbc.ca/news/technology/google-eu-fine-1.4179272">fining</a> Google’s parent company, Alphabet, US$3.6 billion for antitrust violations.</p>
<h2>Another EU probe</h2>
<p>The EU is also investigating the company for requiring other search sites to show Google ads. A third inquiry is examining the firm’s insistence that all Android devices give priority to Google over other search sites. <a href="http://www.businessinsider.com/eu-may-give-google-a-second-record-fine-android-2017-7">More fines</a> could follow.</p>
<p>Google is involved in yet another controversy this week <a href="http://fortune.com/2017/07/11/google-paying-professors-policy-papers/">over its funding to law professors</a> whose research favoured the company’s practices. That funding was not always disclosed.</p>
<p>Google argues that it simply gives consumers what they want. Nonetheless the allegations suggest that Google may have forgotten its onetime “<a href="http://time.com/4060575/alphabet-google-dont-be-evil/">Don’t be evil</a>” motto. It’s worth watching whether this ethical and legal controversy will harm either its <a href="https://theconversation.com/how-brands-turn-customers-into-devoted-followers-78662">brand image</a> or its <a href="https://theconversation.com/memo-to-gordon-gekko-ethics-not-greed-boost-profits-78983">profits</a>.</p><img src="https://counter.theconversation.com/content/80129/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Online search ads are big business. Retailers have to work hard to compete for visibility in Google’s online searches as the company faces trouble in the European Union over its Shopping site.Michael J. Armstrong, Associate professor of operations research, Brock UniversityAnteneh Ayanso, Professor of Information Systems and Director of Centre for Business Analytics, Brock UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/729192017-06-12T11:02:22Z2017-06-12T11:02:22ZIs there structural racism on the internet?<figure><img src="https://images.theconversation.com/files/173174/original/file-20170609-4794-1e86c1w.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Do people use the internet in ways that disadvantage nonwhites?</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-vector/horizontal-vector-illustration-big-number-people-491897284">magic pictures/shutterstock.com</a></span></figcaption></figure><p>The racial inequalities afflicting Americans and our society today are in many ways a result of <a href="http://www.temple.edu/tempress/titles/2089_reg.html">the result of spatial segregation</a>. White people and nonwhite people tend to live in different neighborhoods, go to different schools and have dramatically different economic opportunities based on their race. That physical manifestation of structural racism has been <a href="http://www.penguinrandomhouse.com/books/46131/invisible-man-by-ralph-ellison/9780679732761/">true historically in this country</a>, and is <a href="http://www.penguinrandomhouse.com/books/220290/between-the-world-and-me-by-ta-nehisi-coates/9780812993547/">still the case today</a>.</p>
<p>Today’s internet is built on a similar spatial logic. <a href="http://dx.doi.org/10.1111/soc4.12190">People travel from website to website</a> in search of content in the same way they travel from neighborhood to neighborhood looking for stuff to do and people to hang out with. <a href="http://dx.doi.org/10.1007/978-3-319-39426-8_8">Websites accrue and compound value</a> as visitor traffic and site visibility increases.</p>
<p>But there is a crucial difference: Internet users have – more or less – complete freedom to travel where they choose. Websites can’t see the color of a user’s skin and police incoming traffic in the same way human beings can and do in geographical spaces. Therefore, it’s easy to imagine that the internet’s very structure – the social environments it produces and the new economies it births – might not be racially segregated the way the physical world is.</p>
<p>And yet the internet does appear in fact segregated along racial lines. <a href="http://dx.doi.org/10.1080/1369118X.2016.1206137">My research</a> demonstrates that websites focusing on racial issues are visited less often, and are less visible in search result rankings than sites with different, or broader, focuses. This phenomenon is not based on anything that individual website producers do. Rather, it appears to be a product of how users themselves find and share information online, a process mediated mostly by search engines and, increasingly, social media platforms.</p>
<h2>Exploring online racism</h2>
<p>Words like “racist” and “racism” are loaded terms, primarily because people almost always associate them with individualized moral and cognitive failures. In recent years, though, the American public has become increasingly aware that racism can apply to cultures and societies at large. </p>
<p>My work looks for online analogues of this systemic racism, in which <a href="https://www.routledge.com/Racial-Theories-in-Social-Science-A-Systemic-Racism-Critique/Elias-Feagin/p/book/9781138645226">subtle biases permeate society and culture</a> in ways that yield overwhelming advantages for whites, at the expense of nonwhites. Specifically, I am trying to determine whether the online environment, one completely constructed by humans, systematically produces advantages and disadvantages along racial lines – whether intentionally or inadvertently. </p>
<p>This is a difficult question to approach, but I begin by assuming that today’s technological systems have developed within a culture and society that is systemically and structurally racist. This makes it possible – even likely – that existing biases operate in similar ways online.</p>
<p>In addition, the historical geographical configurations that produced and perpetuated racial inequality provide a useful guide to investigating what systemic racism might look like online. The online landscape, and how people travel through it, are both important factors to understand this picture.</p>
<h2>Understanding online navigation</h2>
<p>First, I wanted to look at the map – how the web itself is structured by website producers. I analyzed what Alexa.com characterizes as the internet’s <a href="http://www.alexa.com/topsites/category/Top/Society/Ethnicity/African/African-American">top 56 African-American sites</a> using a software program called <a href="http://uberlink.com/">Voson</a>. Voson crawls the web to identify what websites the source sites link to, and what sites link to the source sites.</p>
<p>Then I set out to determine the racial content, if any, of each of those thousands of websites, to begin measuring any inequalities that might exist in the online landscape.</p>
<p>Measuring spatial inequality offline typically involves measuring attributes of the people who live in a specific geographic location. For example, ZIP code 65035 designates a “white” neighborhood because <a href="https://factfinder.census.gov/bkmk/table/1.0/en/DEC/10_SF1/QTP3/8600000US65035">99.5 percent of the people residing there</a> (Freeburg, Missouri) are white, according to U.S. census data. By contrast, ZIP code 60619, an area in Chicago, would be considered “nonwhite,” because <a href="https://factfinder.census.gov/bkmk/table/1.0/en/DEC/10_SF1/QTP3/8600000US60619">0.7 percent of its residents are white</a>.</p>
<p>To make this type of distinction between websites, I relied on website metatags – website producers’ descriptions of the site coded to be picked up by and reflected in search engine results. I designated as “racial” websites with metatags including terms such as “african american,” “racism,” “hispanic,” “model minority” and “afro.” Sites without those terms in their metatags I designated “nonracial.” </p>
<p>By using website metatags, I was able to distinguish between racial and nonracial sites (and the segregated traffic between them) based on whether the site’s producers themselves define the site’s identity in racial terms.</p>
<h2>Understanding online navigation</h2>
<p>Once I had labeled each site as racial or nonracial, I looked at the links website producers created between them. There were three possible types of links: between two racial sites, between two nonracial sites, or between a racial site and a nonracial one.</p>
<p>How many of each type of link the data contained would reveal whether bias influenced website producers’ decisions. If there were no bias, the number of links would be proportional to the number of each type of site in the data set. If there were bias, the numbers of links would be disproportionately high or low.</p>
<p><iframe id="SL6kc" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/SL6kc/4/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>While I found slight differences between the ideal theoretical proportions and the actual number of links, they were not significant enough to indicate that any segregation in people’s internet behavior is caused by web producers. People who travel the web just clicking links on websites at random would not arrive at racial or nonracial sites substantially more or less than they should based on the number of such sites that exist. But people don’t just follow links; they exercise their preferences when navigating the web. </p>
<h2>Seeing segregation</h2>
<p>For my second inquiry, I wanted to find out how people actually move between websites. I looked at the same 56 sites as for the previous analysis, but this time used <a href="https://www.similarweb.com/">Similarweb</a>, a prominent web traffic metrics site. For each site, Similarweb produces data showing what websites people came from and what websites people navigated to next. I characterized those sites, too, as “racial” or “nonracial,” and identified three types of paths people took when clicking: between two racial sites, between two nonracial sites, or between a racial site and a nonracial one.</p>
<p><iframe id="UQjbw" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/UQjbw/2/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>In this analysis, the number of clicks between different types of sites would reveal whether bias influenced users’ decisions. I found significantly greater numbers of clicks between nonracial sites, and fewer numbers of clicks between racial and nonracial sites. That indicates that users are going out of their way to visit nonracial sites.</p>
<h2>Capitalizing on search engines</h2>
<p>This gets us closer to the whole story when it comes to segregated traffic patterns and potential inequalities along racial lines. My data also showed that nonracial sites rank significantly higher in search results, and therefore likely enjoy greater visibility, than racial sites. The racial sites are less visible, get less traffic and therefore likely reap fewer benefits from visibility (such as advertising revenue or higher search engine rankings).</p>
<p>It might be tempting to suggest that this merely reflects user preferences. That could be true if users knew what websites they want to go to, and then navigate directly to them. But usually, users don’t. It’s <a href="https://www.emarketer.com/Article/How-Much-Search-Traffic-Actually-Comes-Googling/1011814">much more likely</a> that people type a word or phrase into a search engine like Google. In fact, direct traffic accounts for only about one-third of the traffic flow to the web’s top sites. To quote a <a href="https://www.brightedge.com/sites/default/files/Cracking%20the%20Content%20Code.pdf">conclusion from search optimization firm Brightedge</a>, “overwhelmingly, organic search trumps other traffic generators.”</p>
<p>While more research is of course necessary, my work so far suggests that in conjunction with users’ preferred choices to navigate to nonracial sites more than racial sites, search engines do something with a similar effect: Nonracial sites rank significantly higher than racial sites. That can give racial sites less traffic and less financial support in the form of advertising revenue. </p>
<p>In both of these situations, people and search engines steer traffic in ways that give advantages to nonracial websites and disadvantages to racial sites. This approximates what, in the offline world, is called systemic, structural racism.</p><img src="https://counter.theconversation.com/content/72919/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Charlton McIlwain does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>The physical world is racially segregated as a result of structural racism. A researcher examines whether similar problems exist online.Charlton McIlwain, Associate Professor of Media, Culture, and Communication, New York UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/766882017-05-05T18:03:54Z2017-05-05T18:03:54ZFake news, echo chambers and filter bubbles: Underresearched and overhyped<figure><img src="https://images.theconversation.com/files/167927/original/file-20170504-4929-1sx8gvi.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Don't panic: An international survey finds concerns about fake news are overblown.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-vector/businessman-working-on-laptop-pop-art-356914784">studiostoks/shutterstock.com</a></span></figcaption></figure><p>In the early years of the internet, it was revolutionary to have <a href="https://www.theguardian.com/technology/2014/aug/24/internet-lost-its-way-tim-berners-lee-world-wide-web">a world of information just a click away</a> from anyone, anywhere, anytime. Many hoped this <a href="http://www.hup.harvard.edu/catalog.php?isbn=9780674872332">inherently democratic technology</a> could lead to better-informed citizens more easily <a href="http://www.nature.com/news/society-build-digital-democracy-1.18690">participating in debate, elections and public discourse</a>.</p>
<p>Today, though, many observers are <a href="https://theconversation.com/is-googles-eagerness-to-answer-questions-promoting-more-falsehood-online-70894">concerned that search algorithms</a> and <a href="https://www.theguardian.com/technology/2016/dec/04/google-democracy-truth-internet-search-facebook">social media are undermining</a> the <a href="https://www.theguardian.com/technology/2016/dec/16/google-autocomplete-rightwing-bias-algorithm-political-propaganda">quality of online information</a> people see. They worry that bad information may be <a href="https://www.researchgate.net/publication/220427075_Why_the_Internet_is_Bad_for_Democracy">weakening democracy in the digital age</a>.</p>
<p>The problems include online services <a href="https://theconversation.com/how-can-we-learn-to-reject-fake-news-in-the-digital-world-69706">conveying fake news</a>, splitting users into “<a href="http://www.penguinrandomhouse.com/books/309214/the-filter-bubble-by-eli-pariser/9780143121237/">filter bubbles</a>” of <a href="https://www.wired.com/2016/11/filter-bubble-destroying-democracy/">like-minded people</a> and <a href="http://press.princeton.edu/chapters/s8468.html">enabling users to unwittingly</a> <a href="https://theconversation.com/why-do-we-fall-for-fake-news-69829">lock themselves up</a> in <a href="https://www.psychologytoday.com/blog/psych-unseen/201611/fake-news-echo-chambers-filter-bubbles-survival-guide">virtual echo chambers</a> that <a href="http://www.newstatesman.com/helen-lewis/2015/07/echo-chamber-social-media-luring-left-cosy-delusion-and-dangerous-insularity">reinforce their own biases</a>. </p>
<p>These concerns are much discussed, but have not yet been thoroughly studied. <a href="https://doi.org/10.7717/peerj-cs.38">What research does exist</a> has typically been limited to a single platform, such Twitter or Facebook. Our <a href="http://ssrn.com/abstract=2960697">study of search and politics in seven nations</a> – which surveyed the United States, Britain, France, Germany, Italy, Poland and Spain in January 2017 – found these concerns to be overstated, if not wrong. In fact, many internet users trust search to help them find the best information, check other sources and discover new information in ways that can burst filter bubbles and open echo chambers. </p>
<h2>Surveying internet users</h2>
<p>We sought to learn directly from people about how they used search engines, social media and other sources of information about politics. Through funding from Google, we conducted an <a href="http://ssrn.com/abstract=2960697">online survey of more than 14,000 internet users in seven nations</a>. </p>
<p>We found that the fears surrounding search algorithms and social media are not irrelevant – <a href="https://www.sciencenews.org/blog/science-public/youve-probably-been-tricked-fake-news-and-dont-know-it">there are problems for some users some of the time</a>. However, they are exaggerated, creating unwarranted fears that could lead to inappropriate responses by users, regulators and policymakers. </p>
<h2>The importance of searching</h2>
<p>The survey findings demonstrate the importance of search results over other ways to get information. When people are looking for information, they very often search the internet. Nearly two-thirds of users across our seven nations said they use a search engine to look for news online at least once a day. They view search results as equally accurate and reliable as other key sources, like television news.</p>
<p><iframe id="pfmZV" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/pfmZV/1/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>In line with that general finding, a search engine is the first place internet users go online for information about politics. Moreover, those internet users who are very interested in politics, and who participate in political activities online, are the most likely to use a search engine like Bing or Google to find information online about politics.</p>
<p>But crucially, those same users engaged in search are also very likely to get information about politics on other media, exposing themselves to diverse sources of information, which makes them more likely to encounter diverse viewpoints. Further, we found that people who are interested and involved in politics online are more likely to double-check questionable information they find on the internet and social media, including by searching online for additional sources in ways that will pop filter bubbles and break out of echo chambers.</p>
<h2>Internet-savvy or not?</h2>
<p>It’s not just politically interested people who have these helpful search habits: People who use the internet more often and have more practice searching online do so as well.</p>
<p>That leaves the least politically interested people and the least skilled internet users as most susceptible to fake news, filter bubbles and echo chambers online. These individuals could <a href="http://dx.doi.org/10.1007/978-3-319-40548-3_74">benefit from support</a> and <a href="https://theconversation.com/the-challenge-facing-libraries-in-an-era-of-fake-news-70828">training in digital literacy</a>.</p>
<p>However, for most people, internet searches are critical for checking the reliability and validity of information they come across, whether online, on social media, on traditional media or in everyday conversation. Our research shows that these internet users find search engines useful for checking facts, discovering new information, understanding others’ views on issues, exploring their own views and deciding how to vote.</p>
<h2>International variations</h2>
<p>We found that people in different countries do vary in how much they trust and rely on the internet and searches for information. For example, internet users in Germany, and to a lesser extent those in France and the United Kingdom, are more trusting in TV and radio news, and more skeptical of searches and online information. Internet users in Germany rate the reliability of search engines lower than those in all the other nations, with 44 percent saying search engines are reliable, compared with 50 to 57 percent across the other six countries.</p>
<p><iframe id="nQXkq" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/nQXkq/2/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>In Poland, Italy and Spain, people trust traditional broadcast media less and are more reliant on, and trusting of, internet and searching. Americans are in the middle; there were greater differences within European countries than between Europe as a whole and the U.S. American internet users were so much more likely to consult multiple sources of information that we called them “media omnivores.”</p>
<p>Internet users generally rely on a diverse array of sources for political information. And they display a healthy skepticism, leading them to question information and check facts. <a href="https://www.theguardian.com/technology/2016/nov/29/facebook-fake-news-problem-experts-pitch-ideas-algorithms">Regulating the internet</a>, as some have proposed, could undermine existing trust and introduce new questions about accuracy and bias in search results.</p>
<p>But panic over fake news, echo chambers and filter bubbles is exaggerated, and not supported by the evidence from users across seven countries.</p><img src="https://counter.theconversation.com/content/76688/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>William H. Dutton received funding from Google through Michigan State University to conduct the survey on which these findings are based. However, Google did not design the study, questionnaire, analysis, or findings. The opinions are those of the author and not any organization that supported this research. </span></em></p>Concerns over filter bubbles and fake news are often based on anecdotal evidence. There is relatively little systematic research on the topic; a new survey finds widespread fears are unwarranted.William H. Dutton, Professor of Media and Information Policy, Michigan State UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/739702017-03-09T19:22:56Z2017-03-09T19:22:56ZFaking it: we should make manipulating algorithms for political purposes a crime<figure><img src="https://images.theconversation.com/files/160131/original/image-20170309-21026-1vrdod5.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Can you trust the news you get on social media?</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>There has been a lot of noise about “<a href="https://theconversation.com/au/topics/fake-news-33438">FAKE news</a>” recently, including its ability to <a href="https://www.theguardian.com/politics/2017/feb/26/robert-mercer-breitbart-war-on-media-steve-bannon-donald-trump-nigel-farage">influence the political process</a>. But now some authorities are <a href="https://www.theguardian.com/technology/2017/mar/04/cambridge-analytics-data-brexit-trump?CMP=fb_a-technology_b-gdntech">fighting back</a>. </p>
<p>So far this battle is mainly being played out on the arena of national defence strategies. Little attention has been directed at what the law can do to help combat this destructive force.</p>
<p>We think it would be beneficial to introduce a new criminal offence of algorithmic manipulation for political gain.</p>
<h2>Algorithmic manipulation</h2>
<p>People’s views – including political ones – are subjected to manipulation in a variety of ways. However, we are particularly vulnerable when such manipulation occurs online. </p>
<p>When we’re online, it’s not easy to understand why we see the particular content we are served. Much of what we are shown by search engines and social media is driven by algorithms, and these are largely hidden from view.</p>
<p>Imagine that a marketing company manipulates these algorithms to artificially dictate the search results, tweets and Facebook posts you see, all to communicate content favouring only one brand of product. </p>
<p>Such manipulative and misleading commercial communication is already <a href="http://www.austlii.edu.au/au/legis/cth/consol_act/caca2010265/sch2.html">regulated by law</a>. However, influencing the search results, tweets and Facebook posts you see to favour a particular political view, person or party is not. </p>
<p>This is simply astonishing, because arguably a lot more is at stake in the political than the commercial arena.</p>
<h2>Focus on the means, not the content</h2>
<p>We realise we’re in dangerous territory when regulating non-commercial speech. So if we want to deter certain conduct in this sphere, then any offence we create should be narrow and specific. </p>
<p>Some countries already have in place criminal offences that may be of relevance. Canadian law, for example, contains <a href="http://laws-lois.justice.gc.ca/eng/acts/C-46/section-181.html">the following criminal offence</a>:</p>
<blockquote>
<p>Every one who willfully publishes a statement, tale or news that he knows is false and that causes or is likely to cause injury or mischief to a public interest is guilty of an indictable offence and liable to imprisonment for a term not exceeding two years.</p>
</blockquote>
<p>What we propose is different. We are cautious of provisions such as this as they may be vulnerable to abuse. Instead of focusing on the acceptability of the <em>content</em>, as is done for the mentioned Canadian offence, we turn our attention to the acceptability of the <em>method</em> used to create or facilitate that content. </p>
<p>Focusing on the method is a first way to limit the scope of any offence, while still making it effective in deterring exactly the secretive and invisible manipulation we are concerned about.</p>
<h2>The proposed offence</h2>
<p>We propose the creation of a new offence. So how should we define such an offence? </p>
<p>There are several possibilities, but to keep it simple, and to draw upon the wording used in other settings in the <a href="http://www.austlii.edu.au/au/legis/cth/consol_act/cca1995115/sch1.html">Criminal Code Act 1995 (Cth)</a>, we propose the following:</p>
<blockquote>
<p>A person commits an offence if the person does anything with the intention of dishonestly obtaining a political gain from algorithmic manipulation.</p>
</blockquote>
<p>Terms used need to be carefully defined in the law itself. Thus “political gain” would here include “undermining, or favouring, a political view, organisation or individual”. </p>
<p>We also envisage that a non-exhaustive definition of “algorithmic manipulation” would be useful. Clear evidence of intent would be an essential element of the offences.</p>
<h2>Obvious enforcement difficulties</h2>
<p>We are not naïve about the obvious difficulties involved in enforcing the law we have outlined above. Where the manipulation originates abroad, effective enforcement may require assistance by the very state that is behind the crime in the first place. </p>
<p>And even where the algorithmic manipulation is domestic, proving its existence and identifying the responsible parties will not be easy.</p>
<p>However, these difficulties should not prevent us trying. And law serves different functions, one of them being to communicate societal standards. </p>
<p>If the law is punchy enough, it will have a general deterrent effect. The goal of clearly articulating that algorithmic manipulation for political gain is unacceptable would be achieved the very day such activities are made a criminal offence.</p>
<h2>A delicate balancing</h2>
<p>Any restriction on political activity is a sensitive matter. It goes to the heart of a democratic society. And the last thing we want is to create a tool that may be used to repress a healthy political debate. </p>
<p>However, the question is whether we can <em>retain</em> a healthy political debate in an era of unrestrained algorithmic manipulation. Given recent developments, we fear that we cannot. </p>
<p>As <a href="http://www.nytimes.com/books/first/l/lessig-code.html">Professor Lawrence Lessig famously noted</a> almost 20 years ago: “Left to itself, cyberspace will become a perfect tool of control.”</p>
<p>This seems to be a particularly apt observation when placed in the context of algorithmic manipulation for political gain. We think it is possible to fashion a new offence that specifically and narrowly applies to hidden manipulations of the internet that dishonestly confer political gain, while not deterring free speech and debate.</p><img src="https://counter.theconversation.com/content/73970/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Professor Dan Jerker B. Svantesson held an Australian Research Council Future Fellowship 2012-2016 researching Internet jurisdiction (project number FT120100583). The views expressed herein are those of the author and are not necessarily those of the Australian Research Council. </span></em></p><p class="fine-print"><em><span>William van Caenegem does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>How do you know your search results or social media feeds aren’t being manipulated for political purposes? It’s not a crime to do so. But we believe it should be.Dan Jerker B. Svantesson, Co-Director Centre for Commercial Law, Bond UniversityWilliam van Caenegem, Professor of Law, Bond UniversityLicensed as Creative Commons – attribution, no derivatives.