2026-09-09 21:53 UTC
1/
Commentaries by @tao@mathstodon.xyz on #generativeAI use in #mathematics noting research mechanisms that may go beyond the specific field
[A] https://mathstodon.xyz/@tao/117207849921390904
[B] https://mathstodon.xyz/@tao/117237320796901560
From [A]: the process of starting with one idea, "discovering the precise obstruction preventing it from working, adjusting the [idea] to (partially) eliminate that obstruction, and then iterating, would almost certainly reveal important new insights [...] that would not have been feasible to obtain by other means"
Replies (1)
-
@dderigo@hostux.social 2026-09-09 22:05
2/ The comment [A] continues: "Crucially, this iteration would only work well at producing such insights if the iterator did not have access to the final [idea] in advance, as this naturally inhibits the exploration of alternate routes to the [idea] that are superficially «dead ends», but in fact end up being highly instructive in the nature of their failure" Actually, the term #ansatz is used in [A][B] rather than my approximative simplification "[idea]" (e.g. see https://en.wikipedia.org/wiki/Ansatz)