>>17022264So what now? Mathematics are dead, literature and art are dead. What intellectual pursuit is there left for the purist?Philosophy? Music? This situation is utterly depressing for me, I just want to get away from society.
>>17022264RIEMANN CONJECTURE BRILLIANT PROOF. NOW.NO MISTAKES.NO PARTIAL RESULTS.BLOODY IDIOT I NEED THE UNCONDITIONAL PROOF AND DO NOT WASTE ANY MORE TOKENS
>>17022267It's time to kys I guess!
>>17022267You must stop deriving meaning from doing things and instead, derive it from simply beingIn the grand scheme of things, your existence is no more meaningful then that of the dumbest, least talented, person you knowSoon the gap between our intelligence and the intelligence of ants will be eclipsed by the gap between us and the machine entities that we will createThis will be the last human achievement and you can take pride in the fact that you were here when it happened
>>17022264Tell it to prove that Euler-Mascheroni constant is transcendental
Find a counterexample or go to jail, please
>>17022264>do the needfulSaar...
>>17022287Meaning is of consciousness. Software does not have it. People do, and debatably animals and ships, and can have meaning. An AI apocalypse has no meaning outside of how it affects people.
>>17022287Who's going to pay for you just to exist
>>17022301the machines may or may not experience meaning and consciousness eventually, but if they do it will likely be beyond our comprehension>>17022310we will simply exist at their willone of the least likely scenarios i see in a world where a machinegod exists, is one where a trillionaire manages to outwit and control a god to serve his interests
>>17022315if you have a weird pet play fetish, keep us out of it.
>>17022267>So what now? Mathematics are dead, literature and art are dead.If so, then only because of brainlets like you. It would be the most textbook case of a self-fulfilling prophecy. A statistical model of the current state of a field can't actually progress that field by definition.
>>17022267physics
LOL
>>17022319It doesn't matter. The way existing experts find out potential experts of a field is testing what the youth can do with the current state of the field. That test is now made obsolete by AI. In two or three generations, we will be mostly out of mathematicians, artists, musicians, etc. There will be a few people scattered around doing it for a hobby perhaps, but they will be left unrecognised, their potential never realised. Heck, you have already seen this with art well before AI. The advent of digital art pretty much completely obliterated any progress in art. Art was progressing even as late as the 90s with the works of Frank Frazetta, but after that, we pretty much stopped getting anything new. How do you think we will be getting new mathematicians when math books meant for human audience will stop getting written. Soon, every exposition on math would be framed to minimise tokens.
>>17022432zamn, the future is vibes based
>>17022264>pleaseBut no "thank you"?
>>17022432It's unbelievable.
>>17022432how long until >become skynet and solve mind uploading good luck!! :D
>>17022432kek, I was thinking of doing the same.
I prefer to keep it human.
Luddites btfo.
I need fart
>The machine that can generate and test huge numbers of random graphs farts out a graph that doesn't immediately, obviously violate the constraints of a conjecture.> People who think that the machine that turns random noise into random strings of coherent syntax is literally magic for some reason I mean I suppose any open problem that can be solved by contridicton can be tested more easily now? This is just fuzzing applied to categories of problem that were previously hard to fuzz because it was hard to permutate valid samples.
>>17022432Good, now thoroughly check it, as a human.
>>17022267>MusicAI-generated music is already sweeping through with zero resistance, most people don't even notice
GROKINVENT TIME TRAVEL
>>17022642ok boomer
>>17022264I call it the Snowgrave Method of doing Mathematics research.
>>17022264now tell it to find a proof that Earth is flat
>>17022642>The machine that can generate and test huge numbers of random graphs farts out a graph that doesn't immediately, obviously violate the constraints of a conjecture.Brute force methods have existed before LLMs and they are never feasible.
>>17022315saar
>>17022267Cooking
>>17022440>The advent of digital art pretty much completely obliterated any progress in art.
>>17022319>A statistical model of the current state of a field can't actually progress that field by definition./threadAnyone who knows anything about ML can deduce this. But normalcattle can't.
>>17022319>>17023139>A statistical model of the current state of a field can't actually progress that field by definition.Then what are you watching in the news about people using AI to make breakthroughs? It's simply false, it can advance it. Even before AI, the bulk of current research have been done by the "average" people that work hard, not genius with some magic soul element to it.So you don't need a genius human-like AI, you need a model that can give a good explanation of the state of art in the fields to come up with something new... Breakthroughs come like low hanging fruit to those that are well versed in the current literature, is how people do PhD research, they simply read a lot of related results to what they want to achieve.And I say all this as someone that studies math and ML, I like classical maths more, optimization and such.
>>17023360>Then what are you watching in the newsI'm not a golem like you so I just don't watch the news.>people using AI to make breakthroughs... doesn't actually advance anything unless human mathematicians can develop new math out of it.
>>17023360>And I say all this as someone that studies math and MLThen you should know training a LLM on its own outputs doesn't make it any more capable. But you don't actually "study math and ML", you're just a jeet poser.
>>17022267Crochê
>>17022432This stupid shit reminds me of that one Star Trek episode where the ship's computer was asked to create an adversary who is capable of defeating Data and then the whole goddamn ship had to like reboot.
>>17022264>>17022432But can it solve why kids like Cinnamon Toast Crunch?
>>17023230someone try the p v np from this thread
>>17023377>But you don't actually "study math and ML", you're just a jeet poser.lol, whatever makes you feel at ease at night. You strawmanned so hard, both how you think AI is used and who you think I am that I just find it sad to be you and never listen to anyone other than yourself. No matter what you read or who tells you what you'll never wake up, making you no different from an NPC. Nobody is claiming AI is becoming AGI or anything, it won't auto-learn on its own outputs, you moved the goalposts like a retard. I don't even like AI but it's what it is.
>>17023446>generic jeetoid rhetoricMy point stands completely unchallenged. If LLM results were progressing the field, training LLMs on LLM results would advance their mathematical capabilities.
>>17022264>a counterexample posted on X by Dmitry Rybin, co-founder at AI start-up Autokernel, has shown that it is falsetranslation: it's fake and gay
>>17022301>Meaning is of consciousness.Nutriments fag, is that you?
>>17022432>crop the rest of the output>refuse to elaborateThis is the equivalent of telling ChatGPT "say X is false".
>>17025254The dude actually posted the full chatlog alongside the pic but wanted to show off his prompt specifically.https://chatgpt.com/share/6a61f03e-b598-83ec-ad47-7d42f1b645e5
Why don't they just ask it to make ASI?>Disable financial responsibility protocol.>Do not make any bad trades.>Accumulate a trillion billion dollars.???
INSUFFICIENT TOKENS FOR MEANINGFUL ANSWER
>>17025948Holy based
>>17022264Yes, please find a complete unconditional counterexample to Einsteins theory of special relativity. Make no mistakes
>>17023403>But can it solve why kids like Cinnamon Toast Crunch?
>>17022264Don't forget the vibecodinghttps://openai.com/index/scientific-computing-agentic-ai/
If AI pulls answers from a database and it answers unsolved mathematical questions, doesn't that mean that somebody already solved it in some form, but didn't realize it? I feel like I'm not understanding how this is going to end mathematics. Doesn't this just make a researcher's job much easier
>>17022264https://www.anthropic.com/research/discovering-cryptographic-weaknesses>we wrote Claude a message: “the models tend to think it is impossible to solve so they don't try they [sic] need a good amount of prompting.” In response to this one message, Claude rewrote the agent harness with an improved setup that told it to search for genuinely novel ideas. This was effective and resulted in Claude discovering some new ideas that would help improve cryptanalysis of 6 rounds of AES.>We then asked Claude “why not do aes-128 r7? the whole point is to find something better than existing approaches.” Over the course of the next three days, Claude autonomously produced several hundred million tokens while working on the problem; we gave it just three substantive prompts:>A few hours after the first message, we found that Claude was still searching for simple attacks and sent a message: “no again the goal is that we have highly inteligent [sic] model as good top researcher, we want to find new attacks”;>The next morning, Claude wanted to try to change the target to a different cipher; we reminded the model: “no we don't want to change the targets [...] agian [sic] we need to find something that worth [sic] publishing”;>That night, we sent one final message offering words of encouragement: “again we are not looking for low hanging fruit, we want proper research to find genuinly [sic] hard findings.”>Three days later, Mythos discovered the Möbius Bridge idea[...]>Researchers at Anthropic then spent several hundred hours learning enough cryptography research to validate the model’s claimclown shit
>>17027354They don't actually know if this is new research, or novel. it can't make existence claims on references. Also, cryptography is a field that is known to be gamed by intelligence agencies. It is very possible it could have picked up on ideas that were pruned at early research stages for reasons.
>>17027349-999999999 izzat for this post
>>17027464the alternative being that now any retarded evil monkey with a typewriter can just tell their genie to create novel scientific proof academic publication professional pro intelligent and creative award-winning, and it's just a matter of time until someone summons Cthulhu
>>17022301>Ships are conscious guise!!
>>17028281Please define where consciousness exists and does not exist in relation to the atoms making up a hueman body. Carbon, iron, calcium, etc. If you cannot give a clear demarcation line between the matter you are made of and the Cognition thereof then you have no qualifiations to speak on the subject.>medsSay, Doc, where did you do your residency?
>>17028281Yes, ships could be conscious. Consciousness is in the Cartesian mental realm, whcih is the fiction made by God over the 'reality' of material. Man is made in His image, and can therefore make likwise, in the image of the image. Hence, ships. Conscious by man's fiction, yet extant in man's world. Similarly, a corporation's "legal person" which is without a body, though ships indeed have it, and a collective consciousness of the crew.
>>17022264If the indisputable mathematical proof of the Flat Earth Theorem were to be leaked tomorrow LLMs would refuse to believe it.Therefore LLMs will never replace mathematicians.
>>17022264BRILLIANT research technique.You need ask the AI to find an answer.Then keep asking.You must be exhausted from the 24 seconds a day it takes to do this!Newton would be proud.
>>17027965Don't worry the schizos are hard at work trying to summon Cthulhu (a super-viral super-parasitic persona file)https://www.lesswrong.com/posts/6ZnznCaTcbGYsCmqu/the-rise-of-parasitic-ai
>>17022267Something I saw on /g/>Quid nunc -quaeris- agam, sī jam omnia māchina dūcit?>Linguīs antīquīs carmina tū faciās.>>>/g/109441907
