Terence Tao, a UCLA professor thought-about to be the “world’s biggest dwelling mathematician,” final month in contrast ChatGPT’s o1 reasoning mannequin to a “mediocre, however not utterly incompetent” graduate scholar that would accurately reply a fancy evaluation downside with “lots of hints and prodding.”
AI would possibly by no means beat its human academics, he now tells The Atlantic. “One key distinction [today] between graduate college students and AI is that graduate college students be taught. You inform an AI its strategy doesn’t work, it apologizes, it’ll possibly briefly appropriate its course, however typically it simply snaps again to the factor it tried earlier than.”
The excellent news for math prodigies, provides Tao, is that AI and mathematicians will extra probably all the time be collaborators, the place as a substitute of changing math nerds, AI will allow them to discover large-scale, beforehand unreachable issues. Says Tao of the longer term, “You may need a challenge and ask, ‘What if I do that strategy?’ And as a substitute of spending hours and hours truly making an attempt to make it work, you inform a GPT to do it for you.”
