The GovLab and the Burnes Center for Social Change are excited to announce a new report, Using Artificial Intelligence to Accelerate Collective Intelligence, which introduces a set of practical tools and methods that institutions can use to engage groups in solving public problems using a combination of artificial intelligence (AI) and collective intelligence (CI).
Co-authored with Citizens Foundation, the report describes a novel problem solving method that combines the collective wisdom of human experts and the computational power of AI agents to enhance and scale up public problem-solving processes.
First, we describe Smarter Crowdsourcing – the participatory approach we designed and use with institutions to engage the collective intelligence of human experts in efforts to solve public problems.
Second, we introduce a new AI toolkit, called Policy Synth, which leverages AI agents to automate and scale up the Smarter Crowdsourcing method. By integrating AI capabilities into this participatory process, Policy Synth aims to make the Smarter Crowdsourcing method more efficient, effective, and scalable, enabling institutions to develop solutions more rapidly than would be possible with human intelligence alone.
Finally, we describe a case study comparing solutions to a complex policy problem generated using artificial intelligence to those developed through expert crowdsourcing. The case study demonstrates how generative AI can make participatory problem-solving with human participants more effective because the tools make it possible to translate ideas into implementable proposals more quickly than human participation alone.
Though many AI toolkits exist, they too often are not fitted to the needs of institutions and policymakers. Further, while many existing approaches view AI as a tool to make crowdsourcing and deliberative processes better and more efficient, Policy Synth goes a step further, recognizing that AI can also be used to synthesize the findings from engagements together with research to develop evidence-based solutions and policies.
The report is aimed at policymakers, scholars in the fields of artificial intelligence and collective intelligence, and the civic community, particularly, those who are interested in learning more about the power of open source software, GPT agents, and large language models (LLMs).
Using Artificial Intelligence to Accelerate Collective Intelligence, by Róbert Bjarnason, Dane Gambrell, and Joshua Lanthier-Welch, is available for free download here. Bjarnason and Gambrell will present the report in a workshop at the ACM Collective Intelligence Conference, hosted in June 2024 by Northeastern University’s D’Amore-McKim School of Business.
----------
Our partner project, InnovateUS, also launched a new resource for public problem-solvers yesterday. We are pleased to share the launch of What Works: Fast Field Scanning with AI, a new, free online course for public professionals developed in partnership with Results for America and the Solutions Journalism Network. This at-your-own pace course is designed to equip public professionals with the skills to rapidly identify and implement evidence-based solutions to urgent public problems.
