alt Hacker NewsUnless you claim Tao is outright lying, his statements say he's getting objective value out of AI tools. They are helping him accelerate actual math research.
Math seems like an ideal target for AI, as it provides a firm basis for identifying correctness (has something actually been proved.) I think the consequences of this are going to be enormous for mathematics, including the conversion of the entire historical math literature to formalized, checked proofs. Once that is done, once all that training data is available, the capabilities of math AIs are IMO likely to be extraordinary. Add to that the sort of Alpha Go-style self training as AI provers work on new problems.
> Unless you claim Tao is outright lying, his statements say he's getting objective value out of AI tools. They are helping him accelerate actual math research.
I lack the personal relationship with Tao necessary to determine if his words are truthful or otherwise (namely for me, paid for), but you also failed to provide "his statements" that say either "he's getting objective value out of AI tools" (however one may quantify such a statement) or that "they are helping him accelerate actual math research" choosing instead to simply claim those words as Tao's own.
All of the future tense coded "I think", "going to be", "Once that is done", and "IMO likely" in your second paragraph seem to be at conflict, in the very least tensewise, with your gp comment's past tense of "Not only said, but demonstrated."