Discriminator-generator gap seems to be the most important idea in AI for scientific innovation. With compute + clever search, anything that we can measure will be optimized. First up will be environments that can be verified quickly, with continuous reward, and at scale.

→ View original post on X — @_jasonwei