tag:theconversation.com,2011:/us/topics/collective-intelligence-65640/articlesCollective intelligence – The Conversation2023-08-18T14:44:32Ztag:theconversation.com,2011:article/2116932023-08-18T14:44:32Z2023-08-18T14:44:32ZTeamwork is not always the best way of working – new study<figure><img src="https://images.theconversation.com/files/543442/original/file-20230818-23-cum7z9.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/fatigue-work-home-tired-businessman-neck-1894670614">Girts Ragelis/Shutterstock</a></span></figcaption></figure><p>Throughout the 21st century, teamwork has come to <a href="https://link.springer.com/article/10.1007/s10955-013-0728-6">define</a> the modern work environment. Driven by advances in communication technology, working collaboratively is, as management experts will tell you, how you harness the “<a href="https://link.springer.com/article/10.1007/s12599-010-0114-8">collective intelligence</a>”.</p>
<p>Collective intelligence <a href="https://theconversation.com/what-smart-bees-can-teach-humans-about-collective-intelligence-110656">is often seen</a> as greater than the sum of its parts: superior to the cumulative individual intelligence of the group’s members. Capitalising on it <a href="https://ndg.asc.upenn.edu/wp-content/uploads/2022/10/Centola_2022_TICS_Network_Science_of_Collective_Intelligence.pdf">is said to</a> improve task accuracy (finding better and more correct answers), and enhance task efficiency (finding good answers faster). This in turn leads to quicker and higher quality completion. In other words, when we work together, our performance improves. This has been one of the major factors shaping our modern societies.</p>
<p>At the same time, though, both <a href="https://journals.sagepub.com/doi/10.1177/26339137231156912">research</a> and popular idiom underline the limits inherent to the concept. If “two heads are better than one” suggests the benefits of collaboration, “too many cooks spoil the broth” suggests the opposite. </p>
<p>I led a <a href="https://journals.sagepub.com/doi/10.1177/26339137231156912">recent study</a> looking at whether training and team composition might affect how efficient people are when working together. We found that the benefits of collective intelligence can be outweighed by the cost of having to coordinate between team members.</p>
<figure class="align-center ">
<img alt="People gathered around an office desk." src="https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/543446/original/file-20230818-18-etnnyc.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">People are better able to learn from one another – and communicate – in person.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/focused-arican-american-millennial-female-team-1517113610">Fizkes/Shutterstock</a></span>
</figcaption>
</figure>
<h2>The dynamics of teamwork</h2>
<p>We designed an experimental study using an existing online citizen science project, <a href="https://www.zooniverse.org/projects/zooniverse/wildcam-gorongosa/about/research">Wildcam Gorongosa</a>. Participants analyse webcam photos taken in Gorongosa National Park, Mozambique, to find and identify animal species and behaviour. </p>
<p>We invited 195 members of the public to our lab in Oxford to participate. The experiment comprised two stages: training, then testing, which they did first on their own and then in teams of two. They had five subtasks to complete: detecting the presence of animals; counting how many there were; identifying what they were doing (standing, resting, moving, eating or interacting); specifying whether any young were present; and identifying the animals from 52 possible species (the option of “nothing here” was included, but not “I don’t know”).</p>
<p>We split the participants into two groups. One received targeted training with images similar to the test set. The other received general training with a diverse range of images.</p>
<p>We found the type of training did indeed affect their performance. For those with general training – the “generalists” – efficiency initially improved, but then declined, once they were tested on the specific set of test images. By contrast, those with targeted training – the “experts” – consistently maintained or improved their performance.</p>
<p><strong>How performance changed during the training and testing stages:</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A graphic." src="https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=188&fit=crop&dpr=1 600w, https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=188&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=188&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=236&fit=crop&dpr=1 754w, https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=236&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/543245/original/file-20230817-29-6c4ix9.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=236&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 average change in efficiency tracks the number of correct classifications per minute.</span>
<span class="attribution"><span class="source">Taha Yasseri</span>, <a class="license" href="http://creativecommons.org/licenses/by-nc-nd/4.0/">CC BY-NC-ND</a></span>
</figcaption>
</figure>
<p>To investigate the impact team dynamics would have, we then formed three types of group: these featured either two experts, two generalists, or a mixed pair. </p>
<p>Surprisingly, we found that neither two generalists nor a mixed group performed better than a single generalist working alone. Even two experts working together did not do better than a single expert.</p>
<p><strong>How the groups’ composition affected their efficiency:</strong></p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="A graphic." src="https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=206&fit=crop&dpr=1 600w, https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=206&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=206&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=259&fit=crop&dpr=1 754w, https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=259&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/543250/original/file-20230817-29-mxuqv1.jpeg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=259&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Efficiency varied over time depending on whether the work was carried out by mixed groups, groups of experts, or single experts.</span>
<span class="attribution"><span class="source">Taha Yasseri</span>, <a class="license" href="http://creativecommons.org/licenses/by-nc-nd/4.0/">CC BY-NC-ND</a></span>
</figcaption>
</figure>
<p>We also found that while having an expert in a group improved accuracy for the more complex tasks, it did not improve the group’s efficiency. In other words, the team got more correct answers but took considerably longer to do so. And for simple tasks, there was no improvement in accuracy from having an expert. Ultimately, the time that team members lost in coordinating with each other outweighed the benefit of adding an expert to the group.</p>
<h2>What can we say about the future of work?</h2>
<p>Research has long shown that underperformance in a group is often due to what social psychologists term “<a href="https://www.sciencedirect.com/science/article/abs/pii/003050738490014X">process losses</a>”. The collective intelligence of a team can, for example, be <a href="https://www.pnas.org/doi/abs/10.1073/pnas.1421692112">adversely affected</a> by social biases and what cognitive scientists call “herding” effects, because these can lead to collective decisions being disproportionately influenced by a few members of the group who are less competent yet more confident. </p>
<p>Further, psychologists speak about “<a href="https://psycnet.apa.org/record/1992-05376-001">social loafing</a>” to describe a person performing poorly because they are part of a group – they have the impression that others will do the job without them needing to contribute. When a large number of team members follow this strategy, it can result in the combined efforts of the team being even lower than the sum of individual efforts. </p>
<p>Research also shows the importance of <a href="https://www.nature.com/articles/s41562-018-0518-x">social learning</a> in the context of effective collaborative working, which our study highlights. The experimental method we implemented involved individual training sessions followed immediately by testing the teamwork – this precluded opportunities for people to learn by observing their coworkers’ performance, and therefore one of the advantages of being part of the group during the learning process was eliminated.</p>
<p>The context in which teamwork and collaboration take place matters, as do the tools available for coordination between team members. As internet-based communication technologies are used not only for large-scale voluntary collaborative endeavours, such as citizen science projects, but also for remote working, it is important to recognise the potential effects of different training approaches and team dynamics.</p>
<p>When team members don’t have the chance to observe other workers and reap the advantages of social learning, and when communication is less efficient than face-to-face interactions, the costs and benefits in the teamwork equation can shift. Our research shows that this is even more pronounced when you’re dealing with simpler tasks that don’t require extensive creative problem-solving. Opting to work individually could indeed be a more viable approach.</p>
<p>The dynamics of teamwork – whether in the workplace or in the context of collective action – are complex. While collaboration offers benefits in specific contexts, it is essential to consider the trade-offs between time, accuracy and efficiency. Coordination comes <a href="https://link.springer.com/article/10.1007/s10484-020-09479-8">at a cost</a>.</p><img src="https://counter.theconversation.com/content/211693/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Taha Yasseri receives funding from EPSRC, IRC, and EC. </span></em></p>Collaboration is often promoted as a tool for harnessing collective intelligence. But sometimes working solo is the best option.Taha Yasseri, Associate Professor, School of Sociology; Geary Fellow, Geary Institute for Public Policy, University College DublinLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/2093272023-07-10T20:10:53Z2023-07-10T20:10:53ZHumans set budgets when facing an uncertain future. So do ants <figure><img src="https://images.theconversation.com/files/536396/original/file-20230709-173516-ajhf4y.jpg?ixlib=rb-1.1.0&rect=0%2C0%2C4261%2C3450&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Simon Garnier</span></span></figcaption></figure><p>Imagine you are looking for a parking spot at a crowded event. You find one far from your destination. Do you decide to take it, or invest more time into hunting a better spot which may or may not exist?</p>
<p>You might resolve this decision by “budgeting”: limiting the resources (time) you will spend looking for a better option before settling for the inferior one. This strategy, which allows us to cut our losses when things don’t pan out as we had hoped, is commonly used when we cannot know the payoff of our choices in advance.</p>
<p>Making decisions under uncertainty is a problem we all face. In <a href="https://www.pnas.org/cgi/doi/10.1073/pnas.2216217120">new research</a> published in the Proceedings of the National Academy of Sciences, we show weaver ants (<em>Oecophylla smaragdina</em>) – much like humans – manage it by budgeting their investment into a task with an uncertain payoff. </p>
<h2>Weaver ants bridge gaps with their own bodies</h2>
<p>Weaver ants link their bodies together to form bridge-like structures called “hanging chains”, which they use for crossing gaps encountered along trails. Chains span several times the size of an individual ant and, most strikingly, are self-organized. </p>
<p>This means chains are formed without the help of leaders or external blueprints. Instead, each individual responds solely to its surroundings and local interactions with neighbours. </p>
<p>Understanding self-organization is central to understanding collective behaviour in animal groups – from flocks of birds to insect swarms – and other systems, including human crowds and traffic flow. </p>
<h2>Chains are a gamble</h2>
<p>Building a chain comes at a cost to the colony. Ants in the chain can’t participate in important colony tasks such as defending the nest and foraging. The cost of the chain is proportional to its length: longer chains are more costly, as they keep more ants occupied. </p>
<p>Chains provide a major benefit too: they allow ants to explore areas that would otherwise be inaccessible, which may offer food sources to the colony. Whether an area contains a profitable resource, however, is unknown to the ants until the chain has been completed. </p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="Three photos showing a chain of ants slowly growing downward from one platform to another." src="https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=320&fit=crop&dpr=1 600w, https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=320&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=320&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=402&fit=crop&dpr=1 754w, https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=402&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/536390/original/file-20230709-21-b82382.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=402&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 chain grows as new ants arrive and join the collective attempt to reach the ground below.</span>
<span class="attribution"><span class="source">Daniele Carlesso</span>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>This makes chain-building a gamble. Colonies must invest capital (a number of ants) into forming a chain, which may or may not pay off. </p>
<p>In our study we asked whether, like humans, ants budget their investment into a task when the payoffs are unknown. We expected ants would stop forming chains when the gap to be bridged became too tall, as the cost of the chain would become too great. </p>
<h2>A simple mechanism for a complex decision</h2>
<p>We initially challenged ants to bridge vertical gaps of 25mm, 35mm and 50mm in height. Ants could comfortably form chains within this range, which allowed us to precisely determine the rules they use to build chains. </p>
<p>A detailed analysis of the ants’ behaviour revealed that joining and leaving events happen primarily in the lowest part (1cm) of chains. This indicates that ants are unable to leave their position if one or more individuals start hanging from them. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/bridges-highways-scaffolds-how-the-amazing-engineering-of-army-ants-can-teach-us-to-build-better-158326">Bridges, highways, scaffolds: how the amazing engineering of army ants can teach us to build better</a>
</strong>
</em>
</p>
<hr>
<p>We then found ants decide how long to stay in a chain by visually assessing their distance from the ground below. The closer to the ground, the longer an ant remains in the chain. </p>
<p>Chain formation is thus modulated by a simple rule: each ant remains in the chain for a length of time proportional to her distance to the ground, and remains stuck in place if one or more ants start hanging from her. The ant will then be able to move only if the other ant(s) leave. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/gRsMfqSr134?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">Ants bridging a 50mm gap. Daniele Carlesso.</span></figcaption>
</figure>
<p>Can this rule predict a distance beyond which ants stop forming chains? We answered this question using a mathematical model, which predicted ants should stop forming chains when the gap is taller than 89mm.</p>
<p>To confirm these predictions, we asked ants to form chains over gaps of 110mm – a distance well beyond the threshold predicted by our model. As expected, ants never formed chains over these gaps. </p>
<h2>Tricking ants into investing more</h2>
<p>If ants use vision to assess their distance from the ground, we should be able to trick them into building very long chains (greater than 90mm) by keeping the ground at a constant distance from the bottom of the chain. </p>
<p>We ran an additional experiment where we could lower the platform ants had to reach using a slider. As the chain grew, we lowered the platform, keeping it just out of reach of the ants. Using this apparatus, we tricked ants into forming chains as long as 125mm. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/bees-are-astonishingly-good-at-making-decisions-and-our-computer-model-explains-how-thats-possible-208189">Bees are astonishingly good at making decisions – and our computer model explains how that's possible</a>
</strong>
</em>
</p>
<hr>
<p>Similar to when we set ourselves a time limit for finding parking, ants set a distance limit before giving up. And they do so using a simple rule – remain in the chain for a length of time proportional to your distance to the ground. </p>
<p>Our results reveal how simple rules can guide groups in making adaptive collective decisions in the absence of payoff information. Not only does this help us understand ants – it also provides an algorithm for decision-making in uncertain scenarios, which can be applied in multi-agent artificial systems such as swarm robotics.</p><img src="https://counter.theconversation.com/content/209327/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Daniele Carlesso receives funding from Macquarie University.</span></em></p>Weaver ants organise themselves into bridges to cross gaps and explore new territory – and new research shows this collective behaviour is governed by a surprisingly simple decision-making rule.Daniele Carlesso, PhD Candidate, Macquarie UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1627212021-06-24T13:55:16Z2021-06-24T13:55:16ZPooling society’s collective intelligence helped fight COVID – it must help fight future crises too<figure><img src="https://images.theconversation.com/files/407979/original/file-20210623-23-14h4dsl.jpeg?ixlib=rb-1.1.0&rect=6%2C140%2C4243%2C2440&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/close-miniature-people-social-network-diagram-509324473">Wisiel/Shutterstock</a></span></figcaption></figure><p>A <a href="https://www.gov.uk/government/news/pm-announces-plan-for-global-pandemic-radar">Global Pandemic Radar</a> is to be created to detect new COVID variants and other emerging diseases. Led by the WHO, the project aims to build an international network of surveillance hubs, set up to share data that’ll help us monitor vaccine resistance, track diseases and identify new ones as they emerge.</p>
<p>This is undeniably a good thing. Perhaps more than any event in recent memory, the COVID pandemic has brought home the importance of pooling society’s collective intelligence and finding new ways to share that combined knowledge as quickly as possible.</p>
<p>At its simplest, <a href="https://www.nesta.org.uk/report/future-minds-and-machines/3-what-collective-intelligence/#content">collective intelligence</a> is the enhanced capacity that’s created when diverse groups of people work together, often with the help of <a href="https://theconversation.com/big-data-brings-new-power-to-open-source-intelligence-26554">technology</a>, to mobilise more information, ideas and knowledge to solve a problem. Digital technologies have transformed what can be achieved through collective intelligence in recent years – connecting more of us, <a href="https://www.nesta.org.uk/report/future-minds-and-machines/">augmenting human intelligence</a> with machine intelligence, and helping us to generate new insights from novel sources of data.</p>
<figure>
<iframe src="https://player.vimeo.com/video/391211147" width="500" height="281" frameborder="0" webkitallowfullscreen="" mozallowfullscreen="" allowfullscreen=""></iframe>
</figure>
<p>So what have we learned over the last 18 months of collective intelligence pooling that can inform the Global Pandemic Radar? Building from the COVID crisis, what lessons will help us perfect disease surveillance and respond better to future crises?</p>
<h2>People want to help scientists</h2>
<p>Responding to new and emerging threats requires new methods for filling data and evidence gaps fast. Collective intelligence methods like citizen science have been widely used in the <a href="https://www.tcv.org.uk/wp-content/uploads/2014/11/citizen-science-and-the-environment.pdf">environmental sector</a> for years, but savvy scientists quickly saw the opportunity to deploy these and other approaches to tap into the public’s appetite to contribute to the COVID-19 response. </p>
<p>Before doctors had access to mass community testing or accurate forecasting, for instance, data provided by the public was a valuable early source of information. For example, researchers at King’s College London quickly developed the <a href="https://covid.joinzoe.com/">COVID Zoe symptom tracker app</a>, to which over 4.6 million people have contributed their symptoms since March 2020. This data played a critical role in helping us understand how the virus affects different groups of people, exposing the <a href="https://covid.joinzoe.com/post/the-20-symptoms-of-covid-19-to-watch-out-for">variety</a> of COVID-19 symptoms people have experienced.</p>
<p>Even gamers have played their part behind the scenes. <a href="https://www.eveonline.com/discovery">Project Discovery</a> is described as a citizen science “mini-game”, in which gamers explore outer space while drawing polygons around clusters of cells. The cell populations they trace around are from flow cytometry data that would ordinarily be painstakingly pored over by scientists to see how a COVID infection affects different types of cell. Over 327,000 gamers have taken part since June 2020, saving scientists an estimated <a href="https://www.ccpgames.com/news/2021/eve-onlines-project-discovery-wins-webby-peoples-voice-award-for-public">330 years of research</a>.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/QygAfOksOZk?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>Perhaps more visibly, vaccine development efforts have also been fuelled by volunteers. Over <a href="https://digital.nhs.uk/dashboards/coronavirus-covid-19-vaccine-studies-volunteers-dashboard-uk#dashboard">500,000</a> people signed up to the UK’s COVID vaccine studies volunteer service.</p>
<p>Scientific training and research funding is not usually geared towards public participation and collaboration. That means, despite the potential, the public is typically excluded from participation in scientific research. Changing this might help us shift the dial on preventing the next pandemic and tackling a whole host of our other <a href="https://www.theguardian.com/education/2020/nov/16/the-rise-of-citizen-science-can-the-public-help-solve-our-biggest-problems">complex challenges</a>, such as climate change. </p>
<h2>Making sense of too much data</h2>
<p>Alongside this rise in citizen science, 2020 was also a bumper year for scientific research, seeing a <a href="https://www.google.com/url?q=http://doi.org/10.2139/ssrn.3712813&sa=D&source=editors&ust=1624473166605000&usg=AOvVaw2IyJwAH6kyjG_ZzrzRUFoR">15% increase</a> in paper submissions. <a href="https://reports.dimensions.ai/covid-19/">Over 475,000</a> COVID-related papers and pre-prints have been shared online as of June 2021. </p>
<p>This feverish scientific reporting, especially intense in the field of health and medicine, has raised concerns about quality control. <a href="https://www.nature.com/articles/d41586-020-03564-y">Traditional processes</a> of peer review have come under strain, with papers increasingly released as <a href="https://theconversation.com/how-the-conversation-handles-pre-prints-155063">pre-prints</a>, before they’ve been peer reviewed. Meanwhile, decision-makers face the challenge of finding the most relevant resources in the face of information overload. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-the-conversation-handles-pre-prints-155063">How The Conversation handles pre-prints</a>
</strong>
</em>
</p>
<hr>
<p>The collaborative health evidence database, <a href="https://www.epistemonikos.org/en/about_us/updated_report">Epistemonikos</a>, offers some relief to these challenges. It uses a combination of machine learning algorithms and crowd validation to identify all of the clinical systematic reviews related to the search query entered by the user. </p>
<p>In the past, it was used by policymakers in Chile to accelerate the process of public health legislation. Since 2020, the team behind Epistemonikos has identified <a href="https://app.iloveevidence.com/loves/5e6fdb9669c00e4ac072701d?utm=epdb_en">more than 6,000</a> systematic reviews related to COVID-19 within their database. Highlighting these has helped health professionals and decision-makers find what they’re looking for amid the noise.</p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1347623661356937224"}"></div></p>
<p>It isn’t just scientific research that has proved difficult to make sense of. The flood of data about the pandemic has also required careful collation, seeing as it often comes from multiple sources and is scattered across different websites and open databases, many of which follow different standards and formats. Data about a crisis is only useful if it’s synthesised and presented in ways that decision makers can understand.</p>
<p>One retrospective study showed how <a href="https://link.springer.com/article/10.1007/s13278-021-00723-5">Google searches</a> involving pandemic-related keywords, like “pneumonia”, could have been used to spot the early warning signs of COVID-19 spreading in Europe. The same finding was reached using <a href="https://www.nature.com/articles/s41598-021-81333-1">Twitter</a> data, and could in the future be reached with data from <a href="https://www.thelancet.com/journals/landig/article/PIIS2589-7500(19)30222-5/fulltext">wearable technology</a>. For now, these novel sources of data aren’t integrated into wider surveillance efforts, but doing so could help governments get better at anticipating crises in the future. </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/big-data-can-help-doctors-predict-which-covid-patients-will-become-seriously-ill-153168">Big data can help doctors predict which COVID patients will become seriously ill</a>
</strong>
</em>
</p>
<hr>
<p>In the US, the absence of a publicly available system for aggregating COVID-related data led to the creation of the <a href="https://covidtracking.com/">COVID Tracking Project</a>. A community of over 300 volunteers collected, curated and analysed data sources to produce the most comprehensive public source of information about COVID in the US. Their efforts helped process under-reported data on those in <a href="https://covidtracking.com/nursing-homes-long-term-care-facilities">long-term care</a> and the incidence of COVID organised by <a href="https://covidtracking.com/race">race and ethnicity</a>.</p>
<p>However, another <a href="https://www.wsj.com/articles/data-driven-website-aims-to-help-global-agencies-make-decisions-on-coronavirus-pandemic-11595343601">promising pandemic initiative</a>, the Collective and Augmented Intelligence Against COVID-19 (CAIAC) project, <a href="http://caiac19.org/">failed</a> to get off the ground, despite the support of UNESCO and the Stanford Institute for Human-Centered Artificial Intelligence. The lesson: productively combining human and machine intelligence could help us deal with overwhelming amounts of data, but it isn’t easy. Creating and maintaining new global data infrastructures takes time, effort and significant investment.</p>
<h2>Diversity enhances collective intelligence</h2>
<p>There’s more we can do to properly harness collective intelligence when facing future crises. More data certainly helps, and those who organise that data can help thrust it before key decision makers as quickly as possible. But who makes the decisions matters too.</p>
<p>With the world taken by surprise, it seems that COVID-19 decision making followed the usual modus operandi of excluding the voices of women and minorities. An <a href="https://gh.bmj.com/content/5/10/e003549">analysis</a> of 115 COVID-19 decision-making and expert task forces from 87 countries, including the UK and the US, found that just 3.5% had gender parity in their membership, while 85.2% were majority men. Would the disproportionate impact of COVID-19 on <a href="https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/892376/COVID_stakeholder_engagement_synthesis_beyond_the_data.pdf">black and ethnic minority</a> communities and <a href="https://www.weforum.org/agenda/2020/09/covid-19-gender-inequality-jobs-economy/">women</a> have been as severe if these expert groups had been more diverse? </p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/49-more-likely-to-die-racial-inequalities-of-covid-19-laid-bare-in-study-of-east-london-hospitals-153834">'49% more likely to die' – racial inequalities of COVID-19 laid bare in study of East London hospitals</a>
</strong>
</em>
</p>
<hr>
<p>The collective intelligence literature has long pointed to <a href="https://www.sciencedirect.com/science/article/abs/pii/S0749597816303193">the potential of diversity</a> in problem solving, but these positive effects can only be realised if institutions actively seek out a variety of voices. Without finding better ways to bring diverse perspectives into <a href="https://www.nesta.org.uk/report/how-make-good-group-decisions/">decision making</a>, we’re not going to get too excited about how equally the benefits of the Global Pandemic Radar, and other future efforts to pool data and intelligence, will be felt. </p>
<p>While COVID has elevated AI-enabled modelling to the heart of government decisions, there is still a long way to go before these models are accessible to ordinary people - something which could help diversify decision making. This is where more creative participatory methods, aimed at helping members of the public explore the consequences of policy decisions and collective behaviours, may have a part to play. </p>
<p>The <a href="https://www.scienceathome.org/games/corona-minister-game/about-corona-minister/">Corona Minister</a> game allows people to explore the consequences of different policy interventions as they navigate trade-offs between public health, the economy and civil rights. Elsewhere, researchers in Denmark have created a <a href="https://www.reuters.com/business/healthcare-pharmaceuticals/denmark-trial-uses-virtual-reality-game-boost-covid-vaccinations-2021-06-07/">VR gaming experience</a> where citizens navigate through crowded scenes and try to avoid infection. The aim of the experience is to help participants engage with the complexity of disease spread and the role played by vaccination. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/ZHnZDxfewX8?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>Making progress in how we can effectively think, decide and act together is an area that receives almost no research investment. We think <a href="https://www.nesta.org.uk/report/future-minds-and-machines/">using AI</a> to make the most of the distributed collective intelligence of large, diverse groups is a major frontier for innovation, and a huge opportunity to prepare the population for a future crisis.</p>
<h2>Invest in bottom-up initiatives</h2>
<p>From Ebola to COVID, we’ve learned time and again that crises require both top-down and bottom-up responses. So while the Global Pandemic Radar is a great step forward, governments who are serious about crisis prevention and response need to start supporting the digital and social infrastructures that enable communities to act intelligently themselves. </p>
<p>In 2020, we saw how existing systems of community action were able to pivot quickly to focus on COVID-19. One of them was <a href="http://metasub.org/">MetaSUB</a>, a global project to build microbial portraits of urban transit systems that’s been around since 2015. With a network of volunteers and scientists in over 100 cities, they take regular swabs from trains and escalators, testing the pathogens they find for any markers of antibiotic resistance.</p>
<p>The pandemic saw them quickly set up the <a href="http://metasub.org/projects/#">MetaCOV project</a>, applying their previous methodology to see how microbial samples changed during the pandemic. Their data <a href="https://spectrum.ieee.org/the-human-os/biomedical/devices/swabbing-the-worlds-cities-for-coronavirus">helped show</a> that the longer COVID-19 was on a surface, the less likely it was to make someone sick. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/EM0qE1NkYao?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>Then there’s the <a href="https://www.cmu.edu/news/stories/archives/2017/september/flu-forecasts.html">FluCast forecasting system</a>, which has been tapping into the “<a href="https://www.amazon.co.uk/Wisdom-Crowds-Many-Smarter-Than/dp/0349116059">wisdom of crowds</a>” to predict seasonal flu trends for the US Centres for Disease Control since 2015. The system was swiftly repurposed into <a href="https://delphi.cmu.edu/covidcast/">CovidCast</a> in 2020, which relies on open data sources and the participation of volunteers. CovidCast now offers real-time data across a range of indicators – including mask wearing and COVID-related visits to doctors – to forecast regional spikes in COVID infections and hospitalisations.</p>
<p>That these systems were already present and connected meant they could rapidly be deployed to serve pressing new requirements. Many community-led responses have, of course, emerged to play a vital role without any existing institutional support, such as groups <a href="https://frontline.live/">creating PPE</a> for struggling hospitals, and communities in India and Nepal tracking <a href="https://restofworld.org/2021/pandemic-desperate-indians-turning-online-crowdsourcing-for-help/?mc_cid=4cd2e241a6&mc_eid=ccb32e3378">oxygen supplies</a> and <a href="https://hospitalbedsnepal.com/">hospital bed availability</a>. Many of these new groups should be supported so that they can mobilise swiftly in future emergencies.</p>
<p>On top of that, greater proactive investment, following the lead of organisations like <a href="https://omidyar.com/the-community-infrastructure-fund-for-mutual-aid/">the Omidyar Network</a>, should now be directed towards community infrastructure. And government institutions should acknowledge that it’s currently too difficult for community projects to <a href="https://www.wilsoncenter.org/publication/stitching-together-solution-lessons-open-source-hardware-response-covid-19">connect into institutions</a>. If they’re <a href="https://journals.sagepub.com/doi/pdf/10.1177/0956247817721413">excluded from</a> formal planning, such groups can’t offer their collective intelligence for the collective good. </p>
<h2>Harnessing collective intelligence</h2>
<p>At its best, collective intelligence can help us respond to crises with greater confidence, clarity and cooperation. But we need to start building and reinforcing these schemes and systems now – before the next crisis.</p>
<p>The pandemic has been tough. But it has also thrust our collective intelligence under the spotlight, whether through neighbourhood WhatsApp groups or international scientific research. As we move towards COVID recovery, placing our bets on new initiatives like the Global Pandemic Radar, we must ensure these lessons aren’t forgotten. We must now invest in the combined power of data, technology and people, which will help us avoid the next outbreak and counter society’s next big crisis.</p>
<hr>
<p><em>This article is part of a series on recovering from the pandemic in a way that makes societies more resilient and able to deal with future challenges. It is supported by <a href="https://theconversation.com/uk/topics/resilience-in-recovery-106634">PreventionWeb</a>, a platform from the UN Office for Disaster Risk Reduction. Read more coverage <a href="https://theconversation.com/uk/topics/resilient-recovery-series-106366">here</a>.</em></p><img src="https://counter.theconversation.com/content/162721/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Aleks Berditchevskaia receives funding from the United Nations Development Programme, the Economic and Social Research Council and the UK Humanitarian Innovation Hub. </span></em></p><p class="fine-print"><em><span>Kathy Peach receives funding from the UK Humanitarian Innovation Hub, the United Nations Development Programme, the Omidyar Network, Wellcome and the Patrick J. McGovern Foundation. She is an adviser to the World Economic Forum. Kathy is a Labour party member. </span></em></p>The WHO is creating a Global Pandemic Radar – an example of collective intelligence that must learn lessons from this pandemic.Aleks Berditchevskaia, Principal Researcher, Centre for Collective Intelligence Design, NestaKathy Peach, Director of the Centre for Collective Intelligence Design, NestaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1549842021-02-10T17:57:00Z2021-02-10T17:57:00ZA new intelligence paradigm: how the emerging use of technology can achieve sustainable development (if done responsibly)<figure><img src="https://images.theconversation.com/files/383292/original/file-20210209-15-221zgk.jpg?ixlib=rb-1.1.0&rect=0%2C5%2C1200%2C788&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Activists highlight some of the United Nations' 17 sustainable development goals in Lima, Peru (February 20, 2017).</span> <span class="attribution"><span class="source">Marco Carrasco/Wikipedia</span>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span></figcaption></figure><p>Every new year offers an opportunity for reflection. It is a time to set new goals and revisit old ones. The start of 2021, then, represents a chance to look at successes and failures in meeting the sustainable development goals (SDGs). Since the United Nations General Assembly set these goals six years ago, the SDGs have served as a guide for what the world needs to achieve a <a href="https://www.un.org/sustainabledevelopment/sustainable-development-goals/">“better and more sustainable future for all”</a>.</p>
<p>There has been progress on all 17 goals, which target poverty, health, and inequality around the globe. Nonetheless, the work remains <a href="https://www.devex.com/news/un-forum-offers-sobering-vision-of-sdgs-progress-and-new-virtual-reality-97713">slow and uneven</a>, hampered by low political will, resource constraints and the Covid-19 pandemic. With the agenda’s 2030 deadline looming, the moment is right to ask if new tools and techniques can be used to accelerate progress. New technologies may provide innovative ways to organize human action.</p>
<p>This month, the GovLab and the French Development Agency (AFD) released a report looking at precisely these possibilities. <a href="https://www.afd.fr/en/technology-development-new-intelligence-paradigm-addo-baumann-mcmurren-verhulst-young-zahuranec">“Emerging Uses of Technology for Development: A New Intelligence Paradigm”</a> examines how development practitioners are experimenting with emerging forms of technology to advance development goals. It considers when practitioners might turn to these tools and provides some recommendations to guide their application.</p>
<p>Broadly, the report concludes that experiments with new technologies in development have produced value and offer opportunities for progress. These technologies – which include data intelligence, artificial intelligence, collective intelligence, and embodied intelligence tools – are associated with different prospective benefits and risks. It is essential they be informed by design principles and practical considerations.</p>
<h2>Four intelligences</h2>
<p>The report derives its conclusions from an analysis of dozens of projects around Africa, including Senegal, Tanzania, Uganda. Linking practice and theory, this approach allows us to construct a conceptual framework that helps development practitioners allocate resources and make key decisions based on their specific circumstances. We call this framework the “four intelligences” paradigm; it offers a way to make sense of how new and emerging technologies intersect with the development field. </p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=427&fit=crop&dpr=1 600w, https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=427&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=427&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=537&fit=crop&dpr=1 754w, https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=537&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/383282/original/file-20210209-13-copus8.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=537&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption"></span>
<span class="attribution"><span class="source">Author provided</span>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span>
</figcaption>
</figure>
<p>The four intelligences include:</p>
<ul>
<li><p><strong>Data intelligence</strong>, which covers all those technologies and methodologies that allow for the analysis or visualization of data to support decision-making. Data intelligence encompasses Internet of Things sensors, open data platforms, and <a href="https://datacollaboratives.org/">data collaboratives</a> –structures that allow actors from across sectors to exchange data to create public value. One example we consider is from Senegal, where an international development company formed a data collaborative with the telecom company Orange to estimate the prevalence of illiteracy. The resulting paper <a href="https://www.econstor.eu/handle/10419/130591">identified</a> hot spots of illiteracy and provided a methodology for future work.</p></li>
<li><p><strong>Artificial intelligence</strong> encompasses algorithms intended to mimic human learning and cognition. AI includes both machine learning (algorithms that learn from and improve their behavior through data) and expert models (systems that attempt to mimic the decision-making of a human expert by following predefined rules). While applications of AI remain limited due to resource constraints, they can be useful where data is readily available. For example, in Tanzania, researchers <a href="https://openknowledge.worldbank.org/handle/10986/6610">applied</a> machine-learning methods on accessible satellite data to assess road conditions; the assessments had a 73% accuracy rate, allowing policymakers to potentially identify and prioritize areas in need of road repairs.</p></li>
<li><p><strong>Collective intelligence</strong> uses networked tools to solicit input from groups. Collective intelligence can involve citizen science tools that allow individuals to collaborate and develop knowledge as well as smarter crowdsourcing platforms that allow organizations to engage with experts. It also includes <a href="http://crowd.law/">crowdlaw</a>, networked technologies that enable public deliberation and policy co-creation. In Uganda, for instance, citizen assembly tools <a href="https://oecd-opsi.org/innovations/amplifying-resident-voices-for-better-infrastructure-in-uganda/">have helped</a> citizens pick which infrastructure projects they want funded. The result – in Uganda and elsewhere – is not just a more transparent process but a more legitimate one as well.</p></li>
<li><p>Finally, <strong>embodied intelligence</strong> deploys data and AI in the physical world to automate energy- and time-intensive processes. This form of intelligence often includes expensive tools like unmanned aerial vehicles and 3-D printing, and there are only a handful of examples in the field. However, in Nairobi, some local companies <a href="https://www.france24.com/en/20200519-we-can-get-it-done-here-african-tech-tackles-coronavirus-locally">have used</a> 3-D printing to rapidly create plastic face shields amid the Covid-19 pandemic. One company produces up to 500 face shields per day to support public health authorities.</p></li>
</ul>
<h2>Principles to inform use</h2>
<p>The framework demonstrates the value emerging technologies can bring to development, while also outlining some cautionary thoughts and steps that may be needed to optimize that value. Like any tool, technologies such as AI and crowdsourcing can be used well or poorly, in ways consistent with development goals and ways that are not. It is that critical practitioners approach these technologies as only some options available among many and ask themselves whether and when high-tech solutions are truly preferable to existing methods.</p>
<p>Answering such questions are not easy, but development practitioners can be guided by a few principles. We discuss six such principles in the report, among which three stand out.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=135&fit=crop&dpr=1 600w, https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=135&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=135&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=170&fit=crop&dpr=1 754w, https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=170&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/383281/original/file-20210209-21-anl5q1.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=170&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption"></span>
<span class="attribution"><span class="source">Author provided</span>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span>
</figcaption>
</figure>
<p>First, to ensure that using technology is truly justified, practitioners <strong>must ensure the technology fits the intended purpose</strong>. To do so, practitioners should ensure the specific application or use of technology addresses a clear, well-defined need in a way that resonates with targeted beneficiaries and the local context.</p>
<p>Second, practitioners must <strong>balance benefits against risks</strong>, remaining aware not only of how a project can succeed but also of the many ways it can go wrong. All technology projects carry a risk of failure or unintended consequences. Sometimes it may be justified to take these risks, but development practitioners must be clear-eyed and transparent about the risks, both for themselves and for the intended beneficiaries of their actions. Where possible, assessments of risk versus reward should include inputs from target groups.</p>
<p>Finally, practitioners need to ensure their applications of technology are <strong>feasible within the necessary time frame</strong>. Even in developed economies, emerging technologies can take years to launch at a large scale. Developing countries often don’t have the luxury of time, especially during moments of humanitarian crisis. In these circumstances, practitioners need tools and responses that are deployable immediately.</p>
<p>While emerging technologies are not appropriate in all situations, they offer new opportunities to advance the SDGs. As we start a new year, we encourage development practitioners to consider these technologies alongside existing methods and adopt principles to guide their use.</p>
<hr>
<p><em>The authors would like to thank and acknowledge Andrew Zahuranec, Andrew Young, Dominik Baumann and Juliet McMurren, co-authors of the report.</em></p><img src="https://counter.theconversation.com/content/154984/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Stefaan G. Verhulst a reçu des financements de l'Agence Française de Développement (AFD) pour le report</span></em></p><p class="fine-print"><em><span>Peter Addo ne travaille pas, ne conseille pas, ne possède pas de parts, ne reçoit pas de fonds d'une organisation qui pourrait tirer profit de cet article, et n'a déclaré aucune autre affiliation que son organisme de recherche.</span></em></p>A new report from the GovLab and the French Development Agency (AFD) examines how development practitioners are experimenting with emerging forms of technology to advance development goals.Peter Addo, Responsable du DataLab, Agence française de développement (AFD)Stefaan G. Verhulst, Co-Founder and Chief Research and Development Officer of the Governance Laboratory, New York UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1425192020-07-20T14:31:12Z2020-07-20T14:31:12ZCoronavirus: how the pandemic has exposed AI’s limitations<figure><img src="https://images.theconversation.com/files/348383/original/file-20200720-15-qzedvy.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">AI has its limitations in a crisis.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/futuristic-innovative-corona-covid19-virus-doctor-1697708215">Shutterstock</a></span></figcaption></figure><p>It should have been artificial intelligence’s moment in the sun. With billions of dollars of investment in recent years, AI has been touted as a solution to every conceivable problem. So when the COVID-19 pandemic arrived, a multitude of AI models were immediately put to work.</p>
<p>Some <a href="https://www.ft.com/content/877b8752-6847-11ea-a6ac-9122541af204">hunted for</a> new compounds that could be used to develop a vaccine, or attempted to improve diagnosis. Some tracked the evolution of the disease, or generated predictions for patient outcomes. Some <a href="https://www.imperial.ac.uk/news/196234/covid-19-imperial-researchers-model-likely-impact/">modelled the number</a> of cases expected given different policy choices, or tracked similarities and differences between regions.</p>
<p>The results, to date, have been <a href="https://arxiv.org/pdf/2003.11336.pdf">largely disappointing</a>. Very few of these projects have had any operational impact – hardly living up to the hype or the billions in investment. At the same time, the pandemic highlighted the <a href="https://www.researchgate.net/publication/342183520_Can_Your_AI_Differentiate_Cats_from_Covid-19_Sample_Efficient_Uncertainty_Estimation_for_Deep_Learning_Safety">fragility of many</a> AI models. From entertainment recommendation systems to fraud detection and inventory management – the crisis has <a href="https://www.technologyreview.com/2020/05/11/1001563/covid-pandemic-broken-ai-machine-learning-amazon-retail-fraud-humans-in-the-loop/">seen AI systems go awry</a> as they struggled to adapt to sudden collective shifts in behaviour.</p>
<h2>The unlikely hero</h2>
<p>The unlikely hero emerging from the ashes of this pandemic is instead the crowd. Crowds of scientists around the world <a href="https://nextstrain.org/ncov">sharing data</a> and insights faster than ever before. Crowds of local makers <a href="https://www.fastcompany.com/90489974/the-maker-faire-spirit-is-helping-the-world-tackle-the-covid-19-crisis">manufacturing PPE</a> for hospitals failed by supply chains. Crowds of ordinary people organising through <a href="https://covidmutualaid.org/">mutual aid groups</a> to look after each other.</p>
<p>COVID-19 has reminded us of just how quickly humans can adapt existing knowledge, skills and behaviours to entirely new situations – something that highly-specialised AI systems just can’t do. At least yet.</p>
<p>We now face the daunting challenge of recovering from the worst <a href="https://www.theparliamentaryreview.co.uk/news/uk-economy-suffers-worst-monthly-contraction-on-record-in-april">economic contraction on record</a>, with society’s fault lines and <a href="https://www.theguardian.com/world/2020/jun/02/calls-mount-for-public-inquiry-into-uk-bame-covid-19-death-rate">inequalities more visible</a> than ever. At the same time, another crisis – climate change – looms on the horizon.</p>
<p>At Nesta, we believe that the solution to these complex problems is to bring together the distinct capabilities of both crowd intelligence and machine intelligence to create new systems of “collective intelligence”.</p>
<p>In 2019, we funded <a href="https://www.nesta.org.uk/report/combining-crowds-and-machines/">12 experiments</a> to help advance knowledge on how new combinations of machine and crowd intelligence could help solve pressing social issues. We have much to learn from the <a href="https://www.nesta.org.uk/report/combining-crowds-and-machines/">findings</a> as we begin the task of rebuilding from the devastation of COVID-19.</p>
<p>In one of the experiments, researchers from the Istituto di Scienze e Tecnologie della Cognizione in Rome <a href="https://www.nesta.org.uk/feature/collective-intelligence-grants/multi-agent-systems-enhancing-collective-intelligence/">studied the use</a> of an AI system designed to reduce social biases in collective decision-making. The AI, which held back information from the group members on what others thought early on, encouraged participants to spend more time evaluating the options by themselves. </p>
<p>The system succeeded in reducing the tendency of people to “follow the herd” by failing to hear diverse or minority views, or challenge assumptions – all of which are <a href="https://www.reuters.com/article/us-health-coronavirus-britain-path-speci-idUSKBN21P1VF">criticisms</a> that have been levelled at the British government’s scientific advisory committees throughout the pandemic.</p>
<p>In another experiment, the AI Lab at Brussels University <a href="https://www.nesta.org.uk/feature/collective-intelligence-grants/hybrid-intelligent-solutions-collective-risk-situations/">asked people</a> to delegate decisions to AI “agents” they could choose to represent them. They found that participants were more likely to choose their agents with long-term collective goals in mind, rather than short-term goals that maximised individual benefit.</p>
<figure class="align-center ">
<img alt="Figures representing social distancing" src="https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=401&fit=crop&dpr=1 600w, https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=401&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=401&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/348384/original/file-20200720-63094-1w3d70l.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">The crowd can change its behaviour.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/miniature-people-standing-their-own-circle-1709798407">Shutterstock</a></span>
</figcaption>
</figure>
<p>Making personal sacrifices for the common good is something that humans usually struggle with, though the British public did surprise scientists with its willingness to adopt <a href="https://bylinetimes.com/2020/06/01/the-lost-march-how-the-uk-governments-covid-19-strategy-fell-apart/">new social-distancing behaviours</a> to halt COVID-19. As countries around the world attempt to kickstart their flagging economies, will people be similarly willing to act for the common good and accept the trade-offs needed to cut carbon emissions, too?</p>
<h2>New possibilities</h2>
<p>COVID-19 may have knocked Brexit off the front pages for the last few months, but the UK’s democracy will be tested in the coming months by the need to steer a divided nation through tough choices in the wake of Britain’s departure from the EU and an economic recession.</p>
<p>In a third experiment, a technology company called Unanimous AI partnered with Imperial College, London to run an experiment on a new way of voting, <a href="https://www.nesta.org.uk/feature/collective-intelligence-grants/swarm-ai-approach-optimising-socially-acceptable-outcomes/">using AI algorithms</a> inspired by swarms of bees. Their “swarming” approach allows participants to see consensus emerging during the decision-making process and converge on a decision together in real-time – helping people to find collectively acceptable solutions. People were consistently happier with the results generated through this method of voting than those produced by majority vote.</p>
<p>In each of these experiments, we’ve glimpsed what could be possible <a href="https://www.nesta.org.uk/report/future-minds-and-machines/">if we get the relationship</a> between AI and crowd intelligence right. We’ve also seen how widely held assumptions about the negative effects of artificial intelligence have been challenged. When used carefully, perhaps AI could lead to longer-term thinking and help us confront, rather than entrench, social biases.</p>
<p>Alongside our partners, the Omidyar Network, Wellcome, Cloudera Foundation and UNDP, we are investing in growing the field of collective-intelligence design. As efforts to rebuild our societies after coronavirus begin, we’re calling on others to <a href="https://www.nesta.org.uk/report/future-minds-and-machines/10-recommendations/%23content">join us</a>. We need academic institutions to set up dedicated research programmes, <a href="https://www.nesta.org.uk/project-updates/ai-ci-researchmapping/">more collaboration between disciplines</a>, and investors to launch large-scale funding opportunities for collective intelligence R&D focused on social impact. Our <a href="https://www.nesta.org.uk/report/future-minds-and-machines/10-recommendations/%23content">list of recommendations</a> is the best place to get started.</p>
<p>In the meantime, we’ll continue to experiment with novel combinations of crowd and machine intelligence, including launching the next round of our <a href="https://www.nesta.org.uk/project/collective-intelligence-grants/?offset=9">grants programme</a> this autumn. The world is changing fast and it’s time for the direction of AI development to change, too.</p><img src="https://counter.theconversation.com/content/142519/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Kathy Peach works for Nesta, a global innovation foundation. Nesta's Collective Intelligence Experiments Grants programme receives funding from Wellcome, Omidyar Network and Cloudera Foundation. Nesta's Centre for Collective Intelligence Design also has a partnership with UNDP. She is a member of the Labour party. </span></em></p>As talk of recovery begins, it’s time to rethink the role of artificial intelligence in society.Kathy Peach, Head of the Centre for Collective Intelligence Design, NestaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1365482020-04-20T15:36:04Z2020-04-20T15:36:04ZFive ways collective intelligence can help beat coronavirus in developing countries<figure><img src="https://images.theconversation.com/files/329032/original/file-20200420-152597-kyo1rn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Working together to mitigate the pandemic.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-illustration/coronavirus-disease-covid19-outbreak-microscopic-view-1667357200">Shutterstock</a></span></figcaption></figure><p>The COVID-19 pandemic has so far had the greatest impact in <a href="https://coronavirus.jhu.edu/map.html">developed economies</a> with strong health systems. And the results have been terrifying. But the epicentre of the pandemic could soon shift again – to low and middle-income countries, including those already fragile after years of conflict. <a href="https://www.worldbank.org/en/news/press-release/2020/04/02/world-bank-group-launches-first-operations-for-covid-19-coronavirus-emergency-health-support-strengthening-developing-country-responses">Many</a> are <a href="https://www.ghsindex.org/">woefully unprepared</a>.</p>
<p>Nesta recently highlighted how the pandemic has spurred an incredible <a href="https://www.nesta.org.uk/blog/mobilising-collective-intelligence-tackle-coronavirus-threat/">number of collective intelligence initiatives</a> – combining crowd insight, data and machine intelligence to <a href="https://www.foreignaffairs.com/articles/asia/2020-03-20/how-civic-technology-can-help-stop-pandemic">mitigate the crisis</a>. These range from scientists inviting citizens to help them <a href="https://www.geekwire.com/2020/university-washington-coronavirus-puzzle-game-aims-crowdsource-cure/">design virus-fighting proteins</a>, to DIY biology communities collaborating to design <a href="https://crosscut.com/2020/03/crowdsourcing-against-coronavirus-seattle-biologists-work-diy-test-kit">open-source testing kits</a>.</p>
<p>So as developing countries gear up to tackle COVID-19, here are five ideas for simple collective intelligence projects that might help them.</p>
<h2>1. Mapping demand for medical supplies</h2>
<p>Poor countries with low bargaining power and weak health systems will be further challenged by having to compete with rich ones for supplies of masks, ventilators and other vital equipment. Even knowing what equipment is needed where will be a particular challenge in countries where public health information systems are weak.</p>
<p>In 2009, civil society organisations created <a href="https://stockouts.org/">a tool</a> that enabled researchers and activists to map the available supplies of essential medicines across Africa.</p>
<p>Adapting this idea for COVID-19, frontline workers and emergency responders could use existing technology like <a href="https://www.frontlinesms.com/">Frontline SMS</a> to report on missing or low supplies of key equipment to a common website. This data could then be displayed on a map showing shortage locations.</p>
<p>This would enable governments to see the needs of different health facilities, or even their <a href="https://www.covidcaremap.org/maps/us-healthcare-system-capacity/%233.5/38/-96">existing capacity</a> in real-time detail. It would also help humanitarian agencies, businesses and local manufacturers to respond where supplies are low.</p>
<h2>2. Open-source local production</h2>
<p>During other crises, organisations such as <a href="https://www.fieldready.org/">Field Ready</a> have pioneered the localised production of humanitarian supplies – getting vital equipment quickly and cheaply into conflict zones where traditional logistics have failed. The COVID-19 response could incorporate ideas such as these and tap into the <a href="https://medium.com/@RobertLeeRead/analysis-of-open-source-covid-19-pandemic-ventilator-projects-27acf9075f7e">dynamism</a> of the open-source design and engineering communities. </p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/329034/original/file-20200420-152585-n11p5v.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">Mobilising local makerspaces to produce vital equipment.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/global-spread-novel-coronavirus-pneumatia-medical-1665409711">Shutterstock</a></span>
</figcaption>
</figure>
<p>Governments should consider designating local makerspaces that use tools such as 3D printing as “critical infrastructure”, allowing them to continue to operate during lockdown. Connecting them to hospitals and clinics could provide local level manufacturing capacity that could help with supplying essential items such as personal protective equipment (PPE) and spare parts.</p>
<p>Successful designs could then be scaled up by local manufacturers with greater production capacity. This could be supported by a crowdsourced repository of open-source designs – for example, instructions on how to make <a href="https://www.prusa3d.com/covid19/">face visors</a> or use recycled materials for gowns. A rapid vetting process for each published design to determine its field readiness and level of safety would provide valuable additional information.</p>
<h2>3. Identifying community assets</h2>
<p>There is already <a href="https://www.theguardian.com/world/2020/apr/12/virus-hitting-hardest-modern-equivalent-victorian-slums">emerging evidence</a> from the UK that crowded living conditions accelerate the spread of COVID-19 – and globally, <a href="https://unhabitat.org/sites/default/files/download-manager-files/Slum%20Almanac%202015-2016_PSUP.pdf">up to a billion people</a> live in densely populated slums. In 2018, <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5780711/">researchers in India</a> estimated that an influenza-like respiratory disease would have a 44% higher infection rate among slum dwellers than the rest of the population – even with social distancing. A large factor in this is overcrowding.</p>
<p>For people living in cramped conditions in slums, where many family members share a single room, self-isolating in the home is far more difficult. Alternative measures will be needed. </p>
<p>Repurposing the likes of schools and churches could enable those with COVID-19 symptoms to self-isolate quickly. In China, <a href="https://www.theguardian.com/world/2020/mar/09/how-did-china-get-grips-with-coronavirus-outbreak">stadiums were converted</a> into mass quarantine centres, helping to stem the risk of infection within family groups.</p>
<p>Mapping tools such as <a href="http://openstreetmap.org">Open Street Map</a> could be used to identify the locations of such assets, in partnership with municipalities, businesses and community organisations. Humanitarian Open Street Map is already <a href="https://www.hotosm.org/projects/hot-covid-19-response/">mobilising</a> its volunteer mapping communities, while <a href="http://opencitiesproject.github.io/">Open Cities</a> initiatives have substantial expertise in community mapping for crisis resilience.</p>
<h2>4. Smarter surge response</h2>
<p>Many countries already face chronic shortages of health workers. But 89% of the <a href="https://www.who.int/publications-detail/nursing-report-2020">global nurse shortage</a> is concentrated in low and lower-middle-income countries.</p>
<p>During the HIV/AIDS pandemic and Ebola outbreaks, countries rapidly trained and mobilised community health workers from affected communities. Community health workers could now be vital in helping to track the numbers and symptoms of people with COVID-19.</p>
<p>Modifying <a href="https://covid.joinzoe.com/">symptom-reporting tools</a> for community health workers to use would allow governments and humanitarian agencies to identify potential virus hotspots and deploy surge capacity – the ability to scale up (and down) – most effectively within a country. With a fast-moving pandemic and already-strained resources, governments and humanitarian organisations will need to focus and intensify their collaboration.</p>
<h2>5. Medical hiveminds</h2>
<p>The pace of the COVID-19 pandemic is so fast that the usual process of sharing knowledge via peer-reviewed journal articles is often proving too slow. Instead, doctors have been <a href="https://www.bloomberg.com/news/articles/2020-03-24/covid-19-mysteries-yield-to-doctors-new-weapon-crowd-sourcing">joining specialist discussion groups</a> on social media such as Facebook and Twitter – creating a kind of medical hivemind – to develop answers in real time.</p>
<p>One of these, a Facebook group for registered physicians called the <a href="https://m.facebook.com/groups/148806519631345?view=info">PMG COVID19 Subgroup</a> has over 35,000 members worldwide. There may be a risk that mistakes or misinformation could be amplified with this kind of rapid information sharing and content should always be viewed carefully and critically. But so far it has helped develop new treatment protocols. </p>
<p>For poor countries with few doctors, mobilising the collective intelligence of frontline health professionals and humanitarian agencies across the world may help to speed up the generation and distribution of relevant knowledge. Collective intelligence projects like <a href="https://www.nesta.org.uk/feature/ai-and-collective-intelligence-case-studies/wefarm/">WeFarm</a>, which uses text messaging and machine-learning to match farmers in East Africa with others who can help answer their questions (translated into four languages), could provide a model.</p>
<p>Past pandemics have shown that people with low literacy or proficiency in the main national language tend <a href="https://www.healthaffairs.org/do/10.1377/hblog20200331.77927/full/">not to receive adequate public health information</a>. Tapping into the global hivemind would also accelerate the creation of crowdsourced repositories of commonly-used words associated with the virus in mother tongue and native languages, such as the one created by <a href="https://translatorswithoutborders.org/covid-19">Translators Without Borders</a>.</p>
<p>We know from <a href="https://www.nesta.org.uk/toolkit/collective-intelligence-design-playbook/">experience</a> that humanitarian workers can struggle to absorb new innovations during acute emergency responses. But time is critical and by focusing on repurposing existing tools and <a href="https://covidcourse.thegovlab.org/">tested approaches</a>, we may be able to help stem the next wave of the pandemic.</p><img src="https://counter.theconversation.com/content/136548/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Nesta receives funding from UNDP. Nesta funds Humanitarian Open Street Map and Just One Giant Open Lab through its Collective Intelligence Grants programme. </span></em></p><p class="fine-print"><em><span>Ian Gray consults to Field Ready and Nesta. He has also consulted for Translators Without Borders as a mentor on the Humanitarian Innovation Funds Journey to Scale Programme. </span></em></p>Many low and middle-income countries are unprepared for the COVID-19 pandemic. But innovation could help.Kathy Peach, Head of the Centre for Collective Intelligence Design, NestaIan Gray, PhD candidate, University of ExeterLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1335532020-03-12T14:10:25Z2020-03-12T14:10:25ZCoronavirus: seven ways collective intelligence is tackling the pandemic<figure><img src="https://images.theconversation.com/files/320127/original/file-20200312-111227-1rxsg2z.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">We can deal with this together.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/nurse-holding-test-tube-blood-2019ncov-1624576327">Shutterstock</a></span></figcaption></figure><p>Tackling the emergence of a new global pandemic is a complex task. But collective intelligence is now being used around the world by communities and governments to respond.</p>
<p>At its simplest, <a href="https://www.nesta.org.uk/report/future-minds-and-machines/3-what-collective-intelligence/%23content">collective intelligence</a> is the enhanced capacity created when distributed groups of people work together, often with the help of technology, to mobilise more information, ideas and insights to solve a problem.</p>
<p>Advances in digital technologies have transformed what can be achieved through collective intelligence in recent years – connecting more of us, <a href="https://www.nesta.org.uk/report/future-minds-and-machines/%23content">augmenting human intelligence with machine intelligence</a>, and helping us to generate new insights from novel sources of data. It is particularly suited to addressing fast-evolving, complex global problems such as disease outbreaks.</p>
<p>Here are seven ways it is tackling the coronavirus pandemic:</p>
<h2>1) Predicting and modelling outbreaks</h2>
<p>On the December 31, 2019, health monitoring platform <a href="https://www.wired.com/story/ai-epidemiologist-wuhan-public-health-warnings/">Blue Dot</a> alerted its clients to the outbreak of a flu-like virus in Wuhan, China – nine days before the World Health Organization (WHO) released a statement about it. It then correctly predicted that the virus would jump from Wuhan to Bangkok, Seoul, Taipei and Tokyo.</p>
<p>Blue Dot combines existing data sets to create new insights. Natural language processing, the AI methods that understand and translate human-generated text, and machine learning techniques that learn from large volumes of data, sift through reports of disease outbreaks in animals, news reports in 65 languages, and airline passenger information. It supplements the machine-generated model with human intelligence, <a href="https://twimlai.com/twiml-talk-350-how-ai-predicted-the-coronavirus-outbreak-with-kamran-khan/">drawing on diverse expertise from epidemiologists to veterinarians and ecologists</a> to ensure that its conclusions are valid.</p>
<h2>2) Citizen science</h2>
<p>The BBC carried out a <a href="https://www.telegraph.co.uk/global-health/science-and-disease/deadly-pandemic-could-greatest-threat-national-security/">citizen science project in 2018,</a> which involved members of the public in generating new scientific data about how infections spread. People downloaded an app that monitored their GPS position every hour, and asked them to report who they had encountered or had contact with that day.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/budget-2020-new-uk-chancellor-unveils-30-billion-coronavirus-fightback-but-debt-forecasts-look-optimistic-133461">Budget 2020: new UK chancellor unveils £30 billion coronavirus fightback – but debt forecasts look optimistic</a>
</strong>
</em>
</p>
<hr>
<p>This collective intelligence initiative created a huge wealth of data that helped researchers understand who the super-spreaders are, as well the <a href="https://twitter.com/FryRsquared/status/1234491912775782400">impact of control measures</a> on slowing an outbreak. Although the full data set is still being analysed, researchers have released <a href="https://t.co/ivMS09S9jo?amp=1">data</a> to help with modelling the UK’s response to COVID-19.</p>
<h2>3) Real-time monitoring and information</h2>
<p>Created by a coding academy based on official government data, <a href="https://co.vid19.sg/">Covid-19 SG</a> allows Singapore residents to see every known infection case, the street where the person lives and works, which hospital they got admitted to, the average recovery time and the network connections between infections. Despite concerns about potential privacy infringements, the Singapore government has taken the approach that openness about infections is the best way to help people make decisions and manage anxiety about what is happening.</p>
<p>For dashboard enthusiasts, MIT’s Technology Review has a good round-up of the many <a href="https://www.technologyreview.com/s/615330/best-worst-coronavirus-dashboards/">coronavirus-related dashboards</a> tracking the pandemic.</p>
<h2>4) Social media mining projects</h2>
<p>In early February, <a href="https://www.wired.com/story/how-ai-tracking-coronavirus-outbreak/">Wired reported</a> how researchers at Harvard’s medical school were using citizen-generated data to monitor the progress of the disease. To do this, they mined social media posts and used natural language processing to look for mentions of respiratory problems, and fever in locations where doctors had reported potential cases.</p>
<p>This builds on evidence published in a January <a href="https://journals.lww.com/epidem/Fulltext/2020/01000/Harnessing_Tweets_for_Early_Detection_of_an_Acute.10.aspx">article</a> of the journal Epidemiology that found that hot spots of tweets could be good indicators of how a disease spreads. It remains to be seen how effective these initiatives are, or whether they will succumb to the problems that beset <a href="https://www.wired.com/2015/10/can-learn-epic-failure-google-flu-trends/">Google Flu Trends.</a></p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/320128/original/file-20200312-111242-12o4zoi.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">Bringing the data together.</span>
<span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/coronavirus-sinks-global-stock-exchanges-smartphone-1658460916">Shutterstock</a></span>
</figcaption>
</figure>
<p>The reality of people’s experience of the virus is largely absent from media reporting so far, but the importance of <a href="https://www.ids.ac.uk/opinions/pandemics-social-sciences-are-vital-but-we-must-take-the-next-steps/">social sciences in pandemic preparedness</a> and response is becoming increasingly recognised. We should therefore all tip our hats to the citizens of Wuhan who have been archiving and translating social media data from inside China creating <a href="https://www.technologyreview.com/s/615124/coronavirus-china-wuhan-hong-kong-misinformation-censorship/">chronicles of testimonies</a> of those affected, before they get censored by the government.</p>
<h2>5) Serious games</h2>
<p>To speed up the development of drugs to combat coronavirus, researchers at the University of Washington are calling on scientists and the public to play <a href="https://www.geekwire.com/2020/university-washington-coronavirus-puzzle-game-aims-crowdsource-cure/">an online game</a>.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/is-coronavirus-the-end-of-the-handshake-133185">Is coronavirus the end of the handshake?</a>
</strong>
</em>
</p>
<hr>
<p>The challenge is to build a protein that could block the virus from infiltrating human cells. The game is on <a href="https://fold.it/portal/info/about">Foldit</a>, a 12-year-old <a href="https://www.engr.washington.edu/news/trend/aut10_foldit.html">website</a> which has crowdsourced contributions to important protein research from more than 200,000 registered players worldwide.</p>
<h2>6) Open source test kits</h2>
<p>Responding to concerns about lack of access to testing for COVID-19, Nesta Collective Intelligence grantee <a href="https://app.jogl.io/project/118%23news">Just One Giant Lab</a> is behind an effort to <a href="https://crosscut.com/2020/03/crowdsourcing-against-coronavirus-seattle-biologists-work-diy-test-kit">develop a cheap, quick coronavirus test</a> that can be used anywhere in the world. The initiative is crowdsourcing ideas from do-it-yourself biology communities, with the ambition to open source and share designs so that certified labs can easily produce test kits for their communities.</p>
<h2>7) Sharing knowledge</h2>
<p>In a global crisis, sharing collective intelligence about the virus will be a significant factor in our ability to respond and find new treatments. <a href="https://nextstrain.org/ncov">NextStrain</a> pulls in all the data from labs around the world that are sequencing SARS-CoV-2’s genome, and centralises it in one place for people to see in a genomic tree. This open repository, which is built on GitHub, is helping scientists studying cornavirus’s genomic evolution and enabling tracking of how the virus is passed between people.</p>
<p>Researchers have also been <a href="https://www.nature.com/articles/d41586-020-00660-x.">sharing new findings</a> about the virus’ genomic profile through open source publications and preprint sites such as BioRxiv and Chinaxiv. Paywalls are being temporarily lifted on content related to coronavirus in scientific publications such as BMJ and the <a href="https://www.change.org/p/new-york-times-press-sites-to-stop-blocking-coronavirus-news-from-the-public?original_footer_petition_id=12419401&algorithm=promoted&source_location=petition_footer&grid_position=6&pt=AVBldGl0aW9uACsvOwEAAAAAXmJ6O%252FnU90RmMGYyOGJiNA%253D%253D">public is demanding</a> that major news outlets follow suit.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-chronic-stress-changes-the-brain-and-what-you-can-do-to-reverse-the-damage-133194">How chronic stress changes the brain – and what you can do to reverse the damage</a>
</strong>
</em>
</p>
<hr>
<p>Activists on Reddit have gone one step further and bypassed paywalls to create an <a href="https://www.reddit.com/r/DataHoarder/comments/exdka0/the_coronavirus_papers_unlocked_5352_scientific/">open archive of 5,312 research articles mentioning coronaviruses</a>, citing a “moral imperative” for the research to be openly accessible. Newspeak House is crowdsourcing a <a href="https://coronavirustechhandbook.com/">handbook of tools, tech and data</a> for technologists building things to respond to the coronavirus outbreak.</p>
<p>The World Health Organization (WHO) is also compiling all published research into a <a href="https://www.who.int/emergencies/diseases/novel-coronavirus-2019/global-research-on-novel-coronavirus-2019-ncov.">global database</a>, and making <a href="https://openwho.org/channels/covid-19">learning resources</a> about managing COVID-19 for health professionals and decision makers have been made available on the WHO online learning platform. But they have also been criticised for not replying to comments left on their channels, <a href="https://theunwillingcolonizer.com/2020/03/11/covid-19-a-preliminary-analysis-of-digital-risk-communications/">leaving a vacuum</a> instead of a response to rumours and falseholds.</p>
<p>At Nesta’s <a href="https://www.nesta.org.uk/project/centre-collective-intelligence-design/">Centre for Collective Intelligence Design</a> we’ll keep tracking how collective intelligence is being used during the current crisis, and updating our public online noticeboard of <a href="https://trello.com/b/CdTeRRC6/collective-intelligence-projects">collective intelligence projects</a> as often as we can. Please share any examples you come across in the comments.</p>
<p>By working together and sharing knowledge, we have a better chance of beating the pandemic.</p><img src="https://counter.theconversation.com/content/133553/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Aleks Berditchevskaia is a Senior Researcher at Nesta's Centre for Colletive Intelligence Design which supports Just One Giant Lab (JOGL) through its Collective Intelligence Grants Programme, which is funded in partnership with Omidyar network, Wellcome and Cloudera Foundation.</span></em></p><p class="fine-print"><em><span>Kathy Peach is the Head of Nesta's Centre for Colletive Intelligence Design which supports Just One Giant Lab (JOGL) through its Collective Intelligence Grants Programme, which is funded in partnership with Omidyar network, Wellcome and Cloudera Foundation.</span></em></p>We’re better together.Aleks Berditchevskaia, Senior Researcher, Centre for Collective Intelligence Design, NestaKathy Peach, Head of the Centre for Collective Intelligence Design, NestaLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1106562019-01-29T16:35:04Z2019-01-29T16:35:04ZWhat smart bees can teach humans about collective intelligence<figure><img src="https://images.theconversation.com/files/256108/original/file-20190129-108355-1du2j53.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Not only do bees display collective intelligence, they are also flexible when it comes to making group decisions.</span> <span class="attribution"><a class="source" href="https://www.youtube.com/watch?v=cj5LZBmiQu4">Shutterstock</a>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span></figcaption></figure><p>When it comes to making decisions, most of us are influenced to some degree by other people, whether that’s choosing a restaurant or a political candidate. We want to know what others think before we make that choice.</p>
<p>Humans are social animals. So social that we can rarely be independent of others because of our propensity for copying behaviour and communication – also known as <a href="https://press.princeton.edu/titles/10047.html">social learning</a>. </p>
<p>Humans copy each other every day. You might buy the latest trainers because they’re really popular, even though you have no idea how good quality they are. And then you might share that information, perhaps posting a review on social media. This can induce “smarter” purchasing decisions because usually, if a product is popular, it seems less likely it would be of be poor quality. So sometimes social learning can improve our decision making.</p>
<h2>Learning together</h2>
<p>Our social learning ability has led to extraordinary technological <a href="https://press.princeton.edu/titles/10543.html">success</a>. Advances in modern science and technology, from the smart phone to the <a href="https://www.scientificamerican.com/article/what-exactly-is-the-higgs/">Higgs Boson particle</a>, have been made possible not only by genius innovation, but by humans’ <a href="https://www.nature.com/articles/s41467-018-04494-0">ability to learn from others</a>. So social learning is seen as a source of <a href="https://www.sciencedirect.com/science/article/pii/S0169534709002298">collective intelligence</a> – smart decision making among groups of individuals that improves on the ability of one single person. This can be useful in areas such as management, product development and predicting elections.</p>
<p>However, the opposite can also be true. Crowds can also suffer from collective “madness”, when ineffective or harmful knowledge goes viral due to copying – a phenomenon called <a href="https://www.sciencedirect.com/science/article/pii/S1364661309001703">maladaptive herding</a> – which can trigger things like instability in stock markets. </p>
<p>Why do groups of humans sometimes exhibit collective wisdom and at other times madness? Can we reduce the risk of maladaptive herding and at the same time increase the possibility of collective wisdom?</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/cj5LZBmiQu4?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>Understanding this apparent conflict has been a <a href="http://store.doverpublications.com/0486419568.html">longstanding problem</a> in social science. The key to this puzzle could be the way that individuals use information from others versus information gained from their own trial-and-error problem solving. If people simply copy others without reference to their own experience, any idea – even a bad one – can spread. So how can social learning improve our decision making? Striking the right balance between copying others and relying on personal experience is key. Yet we still need to know exactly what the right balance is. </p>
<h2>Smart flexible bees</h2>
<p>Humans are not the only animals to display collective intelligence. Bees are also well known for their ability to make <a href="http://www.hup.harvard.edu/catalog.php?isbn=9780674953765">accurate collective decisions</a> when they search for foods or new nests. What’s more, bees can avoid maladaptive herding. Bees prevent bad information from becoming viral, although they copy each other through communication and social learning. But how do they do it? </p>
<p>In the early 20th century, Austrian behavioural biologist <a href="https://www.nobelprize.org/prizes/medicine/1973/frisch/facts/">Karl von Frisch</a> found that worker honey bees use a kind of “waggle dance” for communicating with each other. In short, these waggle dances are bee versions of online shopping rating systems. Instead of stars or good reviews, bee ratings are based on the duration of the dance. When a bee finds a good source of food, it dances for a long time. When it finds a poor one, the duration of the dance is short or non-existent. The longer the dance, the more bees follow its suggestion to feed there.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/LU_KD1enR3Q?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>Researchers have <a href="https://link.springer.com/article/10.1007/BF00175101">demonstrated</a> that bee colonies will switch their efforts to a more abundant site, even after foraging is already well underway elsewhere, thus preventing maladaptive herding. Collective flexibility is key. </p>
<h2>Not so flexible humans</h2>
<p>The question is, why can’t human crowds be flexible like bees, especially when both have a similar social information sharing system? To examine this, we developed <a href="https://www.nature.com/articles/s41562-018-0518-x">a mathematical model</a> that was inspired by collective honey bee foraging behaviour.</p>
<p>Two key factors were identified for study: conformity – that is, the extent to which an individual follows the majority opinion; and copying tendency – the extent to which an individual ignores their own personal knowledge and relies solely on following others. </p>
<p>We launched a simple online game as a psychology experiment. Participants had to repeatedly choose one of three slot machines. One slot could drop more money than the others, but players didn’t know which one at the outset.</p>
<p>The mission was to identify the best slot and win as much money as possible. Because many people participated in the same experiment, players could see what other participants were doing in real time. Then they could copy or ignore the choices of the others. </p>
<p>The results revealed that a challenging task elicited greater conformity and the copying increased with group size. This suggests that unlike bees, when large groups are confronted with tough challenges, collective decision-making becomes inflexible, and maladaptive herding behaviour is prominent. The popular slot got more popular because people followed the majority choice, even if it was not actually the winning one.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=428&fit=crop&dpr=1 600w, https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=428&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=428&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=538&fit=crop&dpr=1 754w, https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=538&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/256122/original/file-20190129-39344-1udx7sp.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=538&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption"></span>
<span class="attribution"><span class="license">Author provided</span></span>
</figcaption>
</figure>
<p>The study also showed that humans in groups can be flexible, like bees, when either conformity or copying was low. Players were able to switch to a new and better option when the group size was small or a less challenging version of the task was undertaken. Thanks to the low conformity, there were people willing to explore less popular options, who could eventually find the best one as opposed to the one most chosen. </p>
<p>Our results suggest that we should be more aware of the risk of maladaptive herding when these conditions – large group size and a difficult problem – prevail. We should take account of not just the most popular opinion, but also other minority opinions. In thinking this way, the crowd can avoid maladaptive herding behaviour. This research could inform how collective intelligence is applied to real-world situations, including online shopping and <a href="https://www.predictit.org/">prediction markets</a>. </p>
<p>Stimulating independent thought in individuals may reduce the risk of collective madness. Dividing a group into sub-groups or breaking down a task into small easy steps promotes flexible, yet smart, human “swarm” intelligence. There is much we can learn from the humble bee.</p><img src="https://counter.theconversation.com/content/110656/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Wataru Toyokawa receives funding from Japan Society for the Promotion of Science. </span></em></p>Humans are social animals who like to communicate and copy each other. But key to this collective intelligence is flexibility and a little non-conformity.Wataru Toyokawa, JSPS Research Fellow, School of Biology, University of St AndrewsLicensed as Creative Commons – attribution, no derivatives.