As AI brings about enormous opportunities as well as high risks, we are at a critical moment to shape how AI transforms science and society. New projects at the Michigan Institute for Data and AI in Society (MIDAS) mark its intensified effort to help academia, industry, government and community collaborate with each other to shape artificial intelligence (AI) for the benefits of science and society: an AI policy research initiative, a research co-design exercise, and a national conversation on cross-sector collaboration.
Academia has always been essential in the research and innovation ecosystem: it carries out fundamental research which discovers knowledge and fuels future innovation, it uses research to address societal challenges, and it can provide an independent voice on how technologies including AI should be regulated. However, investment in AI research, the accumulation of AI talent, as well as AI research output, are heavily concentrated in industry. This imbalance has alarmed leaders and researchers across sectors because it could be a barrier to realizing the nation’s full potential for AI-enabled research breakthroughs that will bring about significant societal benefits. A new task that faces us is to build cross-sector collaboration to ensure a balanced and vibrant research ecosystem in the AI landscape.
Earlier this year, with a total of $600K of research resources provided by Microsoft, MIDAS has started a new initiative in its Propelling Original Data Science (PODS) pilot funding program. This initiative will support University of Michigan (U-M) researchers to carry out projects to study AI policy and regulation, understand the impact of AI on society, and develop training for researchers on the ethical use of AI. Four project teams, comprising over a dozen faculty in all, have each already started work to explore a specific aspect of AI policy. Such research will provide much needed insight to policymakers and AI developers and users across sectors.
With this initiative to build momentum, MIDAS is now organizing a national conversation in the form of a workshop to develop a new roadmap of cross-sector collaboration. Also funded by Microsoft, the workshop will gather leaders across sectors to discuss effective information sharing and coordination, understanding and respecting different organizations’ priorities, strategically positioning investments, coordinating talents, and ensuring the oversight of the adoption of new technologies.
Complementary to this visioning activity, MIDAS is also leading a research co-design project to develop a concrete example of cross-sector collaboration, with funding from the National Science Foundation. With AI and database researchers, developers, and end users from U-M, Microsoft, and the City of Detroit, the project aims to build a mechanism to enable better alignment of AI and database research with scientific and government data use, and ensure the trustworthiness of research.
The intensified effort at MIDAS to build cross-sector collaboration is an extension of its data and AI for “social good” effort. It also complements MIDAS efforts to enable the wide and speedy adoption of AI in research in a responsible and ethical manner.
Interested in participating? Please contact Dr. Jing Liu (ljing@umich.edu), MIDAS Executive Director who leads MIDAS effort in the strategic direction of building cross-sector collaboration.