Zoë Brammer & Ankur Vora on How a board game helps Google DeepMind plan for AI in science

What if the best way to prepare for a five-year leap in AI wasn’t a report or a roundtable, but a game? Google DeepMind’s Strategic Foresight team built one: a seven-hour role-play called Science 2030, where scientists and policymakers act out how artificial intelligence might reshape research by the end of the decade. People who’ve played it say it taught them more than any policy brief could.

Zoë Brammer and Ankur Vora, who led the development of the game, explain why they built it and the most surprising findings that came out of it. Among other things, they dive into how fragile people’s trust in science investment really is, and make the case for why the role of human scientists might become even more important as AI takes on more of the work itself.