The top three most inspiring collective intelligence experiments in the UK

At its simplest, ‘collective intelligence’ can be understood as the enhanced capacity that is created when people work together, often with the help of technology, to mobilise a wider range of information, ideas and insights.

Collective intelligence (CI) emerges when these contributions are combined to become more than the sum of their parts for purposes ranging from learning and innovation to decision-making.

CI has been around for a long time, for example, in the 19th Century when it took almost 70 years to crowdsource definitions for the first Oxford English Dictionary, but the rise of new technologies that connect more and more individuals over greater distances to share knowledge and skills has transformed what can be achieved through CI.

CI covers a wide range of participatory methods, including crowdsourcing, open innovation, prediction markets and citizen science.

Some of them rely on competition, while others are built on co-operation; some create a sense of community and teamwork, while others operate on the basis of aggregating individual contributions or microtasks.

Academic research on CI is equally varied and draws on many different disciplines, including social science, behavioural psychology, management studies and computer science.

At the Centre for Collective Intelligence Design we believe that to tackle complex problems we need to mobilise all the resources of intelligence available to us.

That’s why we are working to understand how to best combine the complementary strengths of machine intelligence and collective human intelligence. Part of what we do is to champion what is already happening and support it to grow, through awarding grants and providing support and guidance to teams experimenting with cutting-edge CI methods and approaches.

Together with Wellcome Trust, Cloudera Foundation and Omidyar Network, we created a £500,000 fund for experiments that could generate actionable insight on how to advance collective intelligence to solve social problems.

Drawing on the experiments and insights gathered from the latest round of grantees, here are three inspiring collective intelligence experiments that academics are working on in the UK:

1. Can a swarm of robots interacting with humans facilitate social interaction and help people reach informed consensus?

University of Bristol

As we become more polarised in our views, and challenged by the need to make rapid decisions on emerging issues, it can be hard to ensure that we listen to diverse perspectives.

To overcome this challenge, a team from the University of Bristol built a swarm of 100 small robots and tested whether the robots could help a crowd to reach inclusive and informed consensus by communicating opinion diversity.

Participants could input their responses to a question into the robot, which the robot then displayed for other participants to see before these participants responded to the same question.

The robots were deployed in a shopping centre in Bristol, where members of the public were asked about the most important actions people should take to control climate change.

The results demonstrate how robot swarms can be used to engage people on challenging topics to diffuse and influence opinions, serve as a prompt to launch conversations, and empower introverts to share their opinions.

It has interesting implications for deliberative democracy experts and facilitators, conference organisers, pollsters and anyone looking to design better ways of enabling groups of people to make decisions on polarising topics.

2. Will algorithmically moderating a social network to maintain diversity of opinion improve collective forecasting?

Centre for Cognition, Computation, and Modelling (Birkbeck, University of London)

Birkbeck’s experiment explored how people behave in collective forecasting scenarios, and the role that technology can play in supporting collective deliberation and overcoming polarisation online.

The team developed and tested different rewiring algorithms – programmable rules for manipulating who communicates with whom – as a way of steering deliberating groups towards more accurate forecasts. Participants were tasked with forecasting the probability of ten real-world events occurring, then viewed the forecasts of others and were invited to revise their own forecasts.

Different algorithms were used to influence communication between group members, for example one promoted the most extreme opinions while another promoted the most average opinions. The algorithms measurably improved the quality of the average forecasts made by those groups, and the experiment serves as proof of concept for using rewiring algorithms as tools for influencing the accuracy of collective predictions.

The experiment is relevant for policymakers, alongside anyone working in fields where decisions need to be taken that have an objectively correct answer (e.g. medical diagnostics or intelligence analysis).

3. How do groups make decisions to share limited resources among themselves, and how does behaviour spread across social groups?

University of Nottingham, RMIT University and University of Tasmania

The scarcity of common resources, such as water, is set to become a more pressing and global problem in the future. This experiment explored group behaviour in relation to a common shared resource. The team tested different levels of social connectivity and communication within and between overlapping social groups on an economics experiment platform.

Participants were split into groups and tasked with a resource dilemma – they could individually claim from a common resource, but risked receiving nothing if their collective claims exceeded a threshold. The levels of communication and connectivity between groups were altered to see how this affected coordination. It found that different conditions influenced the ability of participants to manage their common resource without exceeding the threshold.

Communication (strongly) and connectivity (modestly), enhanced sustainable resource use. The findings are relevant for local governments and national policymakers, and emergency response teams and others tasked with coordinating the distribution of limited resources.

A real-world example of where this experiment’s findings could have been beneficial is the recent fuel crisis seen in the UK.

Experiments in collective intelligence design for social impact

These three experiments were funded as part of our second collective intelligence grants programme alongside 12 others – the findings of which are detailed in our latest report.

Watch out for our next grants programme report in early 2022 in which we will share the findings from our third round of grantees.

Funding for experimentation is crucial for accelerating learning in the field, to push the boundaries of existing practice and knowledge.

We hope that the insights gathered from the latest report will inspire more funders, decision-makers and academics to take research in collective intelligence further.

Kathy Peach, Co-Director at Nesta’s Centre for Collective Intelligence Design