Re:How close are computers to automating mathematical reason...
by juped on Saturday August 29, 2020 @12:43PM
(#24317367)
Math is and has always been a form of expressive art, but there's been a vocal constituency since at least the early 20c that claims it's bloodless symbol manipulation. Since computers excel at bloodless symbol manipulation, one would expect them to be programmable to generate proofs of propositions; this is uninteresting to mathematicians unless it has some artistic merit.
[ Parent
| on HN ]
Re:How close are computers to automating mathematical reason...
by orange3xchicken on Saturday August 29, 2020 @03:08PM
(#24318643)
Many mathematicians would disagree with the claim that automated proof checking/generation (see COQ & this presentation by Voevodsky: https://www.ias.edu/ideas/2014/voevodsky-origins) and involved techniques (see the entire field of sat & smt solvers) are "uninteresting & without artistic merit".
Re:How close are computers to automating mathematical reason...
by juped on Monday August 31, 2020 @05:40PM
(#24336205)
Good thing I never claimed that, I just said it was uninteresting to mathematicians unless it has artistic merit!
Re:How close are computers to automating mathematical reason...
by cthonicthulu on Saturday August 29, 2020 @01:16PM
(#24317652)
This is a take that I generally agree with. To be honest, I kinda feel the same way about breathless think-pieces talking about GPT-3.