Strongly agree with this public-goods framing, and the examples: - production of novel datasets - prizes for outcomes with high social value - active experimentation for better public services delivery
And more of this precise agenda-setting work.
This recent @ezraklein piece has a nice bit about getting private AI labs to work on public goods. Perfect time to re-share the @IFP proposal I wrote with @zhengdongwang!
https://www.nytimes.com/2026/05/31/opinion/artificial-intelligence-public-good.html
https://ifp.org/benchmarking-for-breakthroughs/