At Nesta’s new Centre for Collective Intelligence Design, we are searching for insights into how we can make the most of collective human intelligence and machine intelligence to solve our complex social challenges.

As a first step, we are offering grants of up to £20,000 for practical experiments that produce actionable insights into how this can be done well.

We know that new digital technologies such as artificial intelligence can help us to analyse, predict, learn, communicate, and make better decisions faster. They can enable groups to become smarter than the sum of their parts, and mobilise human intelligence at greater scale. But it is not a given. It requires careful design and committed orchestration.

At our inaugural ‘designing collective intelligence’ conference last week, we brought together leading academics and practitioners from a variety of backgrounds to share knowledge about what works and to generate ideas and priorities for research that would help advance the emerging field of collective intelligence design.

Despite much promising practice in collective intelligence - ranging from Zooniverse to Consul, and Duolingo to Patient Innovation, there are still many open questions about the best type of collective intelligence approaches for different problems, and how to make existing efforts even more effective.

Here are just 10 of the many ideas and questions that were raised during the day - thoughts which spanned a range of fields and different aspects of collective intelligence design.

General questions

• How might we create and measure an optimal balance between bringing together the right people (quality) and large numbers of people (quantity) in any collective intelligence effort? Is there an optimal balance?
• How can we better measure the effectiveness of collective intelligence against other approaches?
• Are there archetypes of different collective intelligence models that would enable us to replicate, adapt or scale them to other problems or contexts?

Health

• How might non-clinical datasets (e.g., search data, shopping data) be used to identify people with health issues long before they present in the healthcare system and improve prevention?
• Can we use collective intelligence to make mental health more visible to health service providers? For example, identifying physical health conditions that have a mental health base, or predicting how mental health could be affected by changes to physical health?

Humanitarian aid

• How could combining artificial intelligence with collective human intelligence help us make better predictions of how people will really behave in humanitarian crisis situations, rather than relying on current models which assume people will behave rationally?

Environment

• How can we maximise the potential of secondary data sources (such as holiday snaps on Facebook) for citizen science applications such as biomonitoring, in a way that is ethical and safeguards privacy?
• How can we design collective intelligence efforts so that they don’t end at gathering data (for example crowdsourcing data on air quality), but translate into action more effectively?
• How can we connect different collective intelligence initiatives better to build a more complex, composite picture of habitats and environmental change, for example by combining different citizen science initiatives?

Democracy

• How do pressure groups influence a collective intelligence process in the democracy space?

As this list of questions shows, there is an important need to experiment with new approaches and, crucially, learn what works.

If you think you could help answer one of these questions, or if have your own ideas for applied research or practical experiments that would generate evidence on the best approaches to designing or employing collective intelligence, please check out our new grants.

The deadline for expressions of interest is 9 November. More information can be found in our call for ideas.