Seems to be not a coincidence that what AI is good at is correlated with the backgrounds of AI researchers. Demis was a chess prodigy; Jakub (Chief Scientist) and Mark (CRO) of OpenAI were competitive programmers; many IMO medalists at OpenAI, x-ai. If our world initialized with a different random seed and AI researchers instead came from medicine or environmental science, perhaps AI would first diagnose any disease or build hyper-efficient power facilities, before playing board games and solving AIME problems. Very interesting example of the butterfly effect. Obviously we have also made great strides in AI for medicine, chemistry, etc as well, but frontier model evals often emphasize math and coding
Read 79 replies