Productive generalization - Timothy Gowers

[@danielgagliardi727]

Love Tim Gowers

[@SaveSoilSaveSoil]

Same!

[@Achrononmaster]

@1:25:00 you can't predict a "golden age" that'll end, after which mathematics gets dull. First, you don't know what mathematics AI is limited by, after all, they cannot discover what cannot be programmed, and cannot apply mechanical proof algorithms at infinite speed. As far as we know human insight can leap such "gaps". But also, to a beginner it's incredibly difficult to get up to the frontier, and that'll only get worse, so to push the boundaries of mathematics we will need AI. But likewise for AI, no machine can leapfrog over being trained. So what is the era after Gowers' supposed "golden age"? You just cannot know. It could be an even better age where mathematics gets into seriously weird territory that'll look "indistinguishable from magic" to us. One other possibility is humans turn out to actually be motivated (not as in direct kinetic motion) by a non-physical soul, or whatever (who knows?) That's a conflict with strict physical materialism, and mathematics might have something to say about that, with certain physics impossibility theorems and whatnot. If this is the case then the so-called AI superintelligence might simply be impossible. (The AI nerds hubris these days is hilarious. They lost all humility when the LLM's began doing glorified curve fitting. It's Ptolemy all over again but in ANN computer science.) Which means the robots will be at least as limited as we are, if not more, but in different ways. Always handy to have a fast calculator to hand that you can switch off to save power without moral qualms.

[@NuYiDao]

I just want to know if the next math's breakthrough automatically leads to new creative territory, and if that is always taken up.

[@Achrononmaster]

Without "automatic" in your question there is no answer to that unless you carefully define "breakthrough" and "creativity". Those are not mathematical terms. _With_ "automatic" in your question the answer is clearly no! Mathematicians nor machines do not automatically do things. They need to be either told, motivated, programmed, or forced, none of which are automatic.

[@KuyVonBraun]

We usually gush over Dr Fry so it’s nice to see her gush over her hero 💜

[@BatmanPooping]

No we don’t.

[@davidwilkie9551]

The original Computers were trained people who did hand calculations, and the need for this practice to continue is an illustration of how and why AI is far from being a takeover of Humanity.

[@peterpetigrew2869]

1 + 1 = 2 😊

[@cheeseandonions9558]

nice!

[@yogeshnagpal3671]

Quick maffs

[@jeromejean-charles6163]

Sadly a lot is lost because of an almost invisible pointer on the slides ( should be muche bigger and red may be)!

[@gwebangetforever]

For circular objects, by examining tiny angles, the arc length is surprisingly superior than any surrounding straight lines, even outer limit. So Circumference is beyond 2pi.r.

[@adityadhardwivedi634]

1st to watch, like, and seize the comment box

[@fellowcitizen]

First to contest your primacy

[@readingRoom100]

Spammers.

[@adityadhardwivedi634]

@@readingRoom100 I can assure you gentlemen that I am neither a spammer nor a bot. I can pass for you, A Turing Test, if you wish.😁

[@hoixthegreat8359]

Oxford Mathematics uploading interviews from a Cambridge mathematician....

[@agentsmidt3209]

....on a platform created by UC Berkeley and Stanford students owned by a company created by other Stanford students who had an idea about a search engine operating on a technology envisioned by an Oxford professor on a technology created by MIT in conjunction with the US government whose idea of such technology was influenced by a University of Cambridge graduate...etc...

[@cheeseandonions9558]

Such things cannot be!

[@hayekianman]

the sunnis and shias cant get together? ok

[@mathematics5573]

The reason A levels students do 4 A levels now, and not 3, is because the exams are easier than they used to be. That is not good. Do less and learn more maths