liebe

[liebe @liebeix]

@scottnarmstrong I would adapt this old quote to the present by replacing “quite good” with “very good”. But I think the main message of the quote still holds. If LLMs reach a point where they will be able to provide original good ideas then it is not just math, all disciplines will be affected.

[liebe @liebeix]

@Mascobot It’s just hard to imagine what “AI can do them all” means in theory, let alone in practice.

[liebe @liebeix]

@predict_addict @KarlLandheer Yeah but then you should differentiate. You have to make sure grads from top schools stay in your country after their phds.

[liebe @liebeix]

@lianda_edu This is not a good thing

[liebe @liebeix]

@lacker Math research will be still managed by humans. Mathematicians will have lots of work to do.

[liebe @liebeix]

@s_batzoglou I don’t agree with the wording. Math has not “fallen” to AI. Mathematicians will keep doing math, they will just be focusing on different aspects of it.

[liebe @liebeix]

@hayou_soufiane Dominating mathematical discovery and generating proofs are two different things. Machines for now have proved themselves to be really good at the latter. I am not sure about the first one. I wonder on what basis you are saying that they will be “dominating”?

[liebe @liebeix]

@HaghifamMahdi @hayou_soufiane Prestige is not a bad thing.

[liebe @liebeix]

@DavidTurturean lol at this point let’s open a predictions market on this. Or maybe there is already one.

[liebe @liebeix]

@LocasaleLab Well PhD has always been an “over-qualification”. But majority of phds are not hired because of their knowledge. They are hired because they proved that they can think deeply. Especially if you come from a top school…

[liebe @liebeix]

@yoavgo I think the reason is that math never really had a “technological revolution”.

[liebe @liebeix]

@JoergStoye @ObhishekSaha Well in math depends. Sometimes people proved big conjectures and they did not want the paper to be public until reviewers accept it. This is still happening today.

[liebe @liebeix]

@quasicoh Archaic? What? You know that math is many things not just the niche topic you have in mind?

[liebe @liebeix]

@plumsirawit AI may be better than you at solving problems but that’s not what math is really about no? I agree that there will be some kind of revolution in math. Mathematicians will also adapt.

[liebe @liebeix]

@carney It’s pretty simple. Some mathematicians spent their life solving problems that now AI can do. Having said this there are many other mathematicians that are not “problem solvers” but more “theory builders/methodologists” that are now much more productive with ai.

[liebe @liebeix]

@JFPuget Well sure I made a claim you didn’t make 😂
How am I supposed to reply otherwise? With a claim you have made? Not much of a reply isn’t it?
You claimed that you wouldnt purse a PhD in math/cs because AI is replacing us. My reply makes an opposite claim. Replies can also disagree.

[liebe @liebeix]

@DimitrisPapail What math is that?

[liebe @liebeix]

@JFPuget And in this future where math evolves, don’t you see humans involved actively in it? I agree that current conjecture problem solving will be outdated. Humans have always adapted and I don’t think they will want to stop thinking about the world. Math PhDs will also evolve with it.

[liebe @liebeix]

@ricardoalatorre @taobanker Yea only issue now with AI is that solving conjectures has mostly lost value. Somehow we need to redefine what resume credibility means