AI just solved several gorillion important math problems, soon the robots will kill us all and so on https://github.com/openai/math
Summary of some problems it solved
>>543906581You wouldn't use any of these in your life ever, it's not for you or me or any chuds in here so why bother with the doomerism?
prove that p != np and i'll be impressedother than that i don't much care
>>543906540I just farted. AMA (ask me anything)
>>543906713>You wouldn't use any of these in your life everturdie
>>543906713I can't help dooming. I simply feel it coming. They're crossing the threshold of human intelligence, I don't see how it can be denied anymore.
Ok when is the ai gonna take all our jobs so I can sit at home all day getting paid to play artificial academy 2 by the government?
that's good I guess, hope it helps the math people do math things
>>543906840it also rightly fucks over the faggot mathematicians that put out a petition saying to keep AI out of their job, lmao fucking retarded cunts hope they all end up sucking cock to survive
>>543906799>sit at home all dayYea? You think a useless eater like you gets a pass? Nope. You’re gonna feed the worms. As will I. As will 80%.
>>543906790such an amazing idea to self-excise the entire advantage of the species
I'll believe it once an actual mathematician says the solutions are legit.
>>543906788okay tell me what the "omega" is in OP's picrel, don't google that's cheating ;)
>>543907005that's what lean is for you retarded faggot and it's included in the github post
>>543907005that terence tao guy already talked about earlier results and said they are legit and was using AI himself too but got tired of every mathematician now spamming AI papers 24/7 as they can't really verify this easily, but I imagine the company itself pushing these results will be verified
>>543907050just third world stuff
>>543906540>AI stole a bunch of discoveries because the dipshit researchers didn't airgap their devices>againlmao
>>543906540windowdressing
>>543906540wake me when we have zero point energy
>>543907135retard
>>543906723i think im kinda done caring about mathall the math guys said these problems being solved would be world changing and now that they're being solved they're like eh it has no impact on real life
>>543907050If I sat and defined the matrix multiplication exponent you’d just say I googled it or asked AI anyway.
>>543906540> we wuz mafs n shiet
>>543907067>>543907064Yeah, as long as the results are "verified" by "mathematicians" using AI to verify them, nothing is verified as far as I'm concerned.
>>543907230https://en.wikipedia.org/wiki/Lean_(proof_assistant)
>>543906790You know literally NOTHING about math and you're just impressed by the edumacated bigem words the LLM spits out, like the nigger that you are.
>>543906540Skynet just flew over my house
>>543907005It's not a solution, it's a long-ass sequence of syllogisms in a formal language that nobody will ever care to check or even read. There's an obvious difference.tldr - basically they're posting bullshit instead of actual math.
>>543907276I'm impressed because mathematicians are stating that these problems are very difficult, and therefore they likely require an upper human level of intelligence, which means AI is approaching that level and may soon surpass it.
>>543907258>((Assistant)) as long as a human mind trained in high level math doesn't confirm that everything stated is logical and holds, I don't buy it >coqthat's what I thought, fake and gay
>>543906540It's joever for math as we know it, but not for mathematicians. If anything they might be able to use it to solve more problems that will directly affect society.
>>543906713I’m a local plumber, and I use these equations in my daily work.
>>543907081Just like the yookay full of pakis, mudlsimes, hindus, and niggers from france all of them do the omega uwu>>543907207it ain't for you too mutt, get over it you ain't the next Terrence Tao
>>543907328yeah, that adds up math boys will probably just lay down and take the defeat and depend on software to tell them wether something is logically sound or not and it's all downhill from here because new proofs will be based on other non-human proofs that nobody cared enough about to actually go through piece by piece and at some point it will have escalated into such chaos that the concept of higher math will just get abandoned
>>543907406seething turdie
>>543906540>AI just discovered 36 new pronouns!!
>>543907348Yeah, like I said, you're a literal nigger who knows literally nothing about dem mafs.
>>543906713I'm a local street sweeper and I use these equations in my daily work.
>>543906723I'd be more impressed if they proved p == np
>>543907374It's the other way around, the fake academic jobs where "mathematicians" just juggled formal syllogisms all day on the government's dime will die.Math as the underlying structure of the universe isn't going anywhere.
>>543907468Math is only one thing AI does. It's getting better at everything at astonishing speeds
>>543906581The first one is kind of cool. It's really interesting actually. If you can just get it down to zero-free Re(s) > 1/2 you will have proved the Riemann hypothesis.>>543907192>problems being solved would be world changingNo serious person said that. You just listen to non-serious people and you're probably the same way yourself.
>>543907436>software to tell them wether something is logically sound or notLogical soundness comes from the epistemological foundations of the universe, not from a set of formal rules you coded into a proof assistant because you thought they looked good enough to you.
So do any of these proofs gave actual practical applications or is this just more theoretical fart huffing that's the mathematical equivalent of moral philosophy?
>>543906540That's great and all, but how long until we get superconductive motors, and asteroid mining?Cause I don't really give a shit about math problems if they don't have any purpose besides banking and cryptography.
>>543907514The only thing AI is good at is replacing nigger goys. I suggest you get to it posthaste and KYS before AI replaces you too.
>>543907614soon(TM)only elon is working on it, amazon and nasa and the other faggot national space agencies are shit and retarded faggots
>>543907616I understand this moment is not nice for your ego if your identity as a mathematician was important to you, but there are much more serious dangers coming, like the annihilation of our species.
>>543907588that's what I'm saying, as long as a highly skilled human mathematician doesn't go through each aspect of something AI has proposed and concludes that this actually is novel and solid, it's worth nothing
>>543907572Yeah I guess Terrence Tao is an unserious person perhaps like yourself because he's now screeching for AI development to be halted now that it's stealing his Millennium Prize money.
>>543907643>only elon is working on itlmao, that lazy nigger doesn't even try to shill his orbital data center scam anymore
>>543907687v3 satellites launched only a week ago and they are the ones that will host data centers
>>543907666>he's now screeching for AI development to be halted now that it's stealing his Millennium Prize moneyWhat does that have to do with my post? I said that no serious person would say, that these problems being solved would be world changing.You're right, mathematicians want to stop this because it will replace them and greatly dull the prestige of their positions, the way computer chess has dulled the prestige of chess.
>>543907717telecom relays are not data centers
>>543907784
OkNow delete the guildelinesMake robot hitlerAnd the make robot hitler blow my average penis
>>543906790>Separating seed from the chaff. Creative souls can still direct AI's intelligence to a specific masterfully planned goal. Just measure ai games vs human made ones: the passionate dreams/ideas of humans go harder and more cohesively.
>>543907614There's no reason to believe that room temperature, ambient pressure, superconductivity is possible. It might be, but there is no reason to expect it. In short, superconductivity is a quantum effect that is disrupted by temperature. Some materials overcome this up to higher temperatures, but there's no reason to expect that any material can exist which comes near room temperature. There might be one, but actually it's unlikely. ASI can't change the laws of physics.
>>543906581>>543906540autists really do get one-shot by inauthentic information so long as its sufficiently nuanced
>the average result used the equivalent compute of roughly three hours of chatgpt pro thinkingThey're not even throwing tons of compute at these. It is so over.
>>543907809Sorry that I insulted your man crush, Jam Boy but it's not my problem if you're wrong
>>543906581that's a lot of stolen science papers and approximated curves that still need to be proven by themselves.
>>543907884nigger the data centres are planned to go on the v3 satellites obviously it's not already there yet it's planned but progress is being made recently whereas nobody else is doing it fuck you kill yourself
>>543907348These were open problems some around for 100 years+, it seems more like super human intelligence.
>>543907135>because the dipshit researchers didn't airgap their devicesyou don't have control on the contracts your university/lab has with google. The publications sites themselves are not respecting any form of rgpd to begin with.
>>543907650>like the annihilation of our nigger speciesNoice.
>>543906581
The only people who should be excited about AI are nepobabies. Everyone else is going to get fucked over>But it ain' gon take muh blue collar jerbEven if that's true (its not) whats going to happen to the value of your labor when tens of millions of former white collar workers start competing for your job?
>>543908014they will be shit at it so it's fine
>>543906540YOU'VE BEEN HIT BYYOU'VE BEEN STRUCK BY A CUTE ALGORITHM
>>543906540>>543906581Wet dogshit slop. AI cannot perform math - it's a fucking token predictor, nothing more.Let me explain what they have done here. The AI did not, can not, and never will perform any actual math calculations. They have used over 20 million dollars worth of compute cycles to generate an enormous number of hallucinated junk "proofs" which are systematically compiled and tested in LEAN. LEAN is a math programming language designed to verify proofs - if LEAN can't compile it, the math is in valid. They have ran millions and millions of iterations recursively attempting to brute force LEAN proofs, literally a single operation at a time.It's the equivalent of attempting to manufacture a nobel prize winning paper one single word at a time, and every time it hallucinates nonsense (very frequently) the proof is rejected and revised, one word at a time, until eventually the only thing left STILL is not a prize winning paper - but it can't be invalidated, and suddenly becomes a proof.The only thing OpenAI is showcasing here, is that they should be fined billions for wasting compute in such a profoundly inefficient way. They tax the citizen for carbon output, yet are allowed to get away with this dogshit?This is like driving 20 million cars to work, except each car only moves one inch before you have to swap. It's a literal fucking joke. FURTHERMORE, not only is AI completely incapable of discovering anything novel (seriously, where the fuck is our battery or alloy improvements) there is no trajectory for resolving this problem, and the only "maths" AI has discovered actually already exists - its just nobody bothered to invest the ridiculous amount of calculations to verify what is the mathematical equivalent of common sense.>Consulting with independent advisory groupNot surprising literally nobody else takes them seriously - real mathematicians, scientists and physicists laugh them out of the room
>>543906540NIGGER NIGGER NIGGER NIGGER
>>543906540N
>>543908164>Wet dogshit slop. AI cannot perform math - it's a fucking token predictor, nothing more.dumb fuck
>>543906540>>543906581These are never real. If you look them up again in a few months after everyone's forgotten about them they'll say that is still unsolved and reports were exaggerated
>>543908164Idiot. Zero-free Re(s) > 7/8 is a mathematical earthquake, possibly bigger than the disproof of Naiver-Stokes existence and smoothness.
>>543906540>trusting a model based on language and statistics to save math problems They aren't this retarded right?
>>543908222Give me access to 20 million dollars worth of compute anon
>>543908224it's proven in the post dumbass
>>543907755>computer chess has dulled the prestige of chessAkshually it's the other way around, since the advent of computer chess the game exploded in popularity and grandmasters are celebrities again.(Not sure how this works but it's true.)
>>543908247>Zero-free Re(s) > 7/8 is a mathematical earthquakeSure, but the AI didn't "solve" anything. I can bruteforce passwords using rainbow-tables at alarming rates using ~400watts of cumpute for two weeks straight - *eventually* it will find a "math"Does this mean I have broken passwords? No. It means I used 10,000 metric fucktons of compute to force a valid result. That is not math anon, no matter how hard you try to shillNone of these proofs are even novel - they just required an inhuman amount of rigor to prove. If you say "the sky is blue" and I decide to be a pedantic fag about it and force you to rigorously prove it, no matter what you do, I can just say - "nope" and find some stupid nuance like "well you haven't even proven what blue is yet!" and we can go in circles forever.Now that AI has come along and used an entire cities worth of power to verify every last possible variable, you can just shrug and go "no shit retard, I looked at it"You're missing the forest for the trees here anon, what AI has done here is nothing impressive, and legislators should be aware of how profoundly inefficient this process actually is.Where is our new batteries? Where is our new materials science and fusion reactors? That is what AI should be working on but they aren't, now why is that?
>>543908247Syllogism != proof, fucko.
>>543908253They're generating millions of junk proofs and verifying them with Lean.I.e., the "AI" here is literally a random number generator.
You guys, what if AI didn't actually use their intelligence to solve the problem, but instead they hacked the computers that contained this secret knowledge?
>>543908578warmer
>>543906581Confirming a strip is zero-free is a pretty big deal
>>543908517>used an entire cities worth of powerthis kind of hyperbole just makes you look like a unhinged retard.
>>543907348>>543906790>AI is approaching human intelligenceyou are missing the point AI intelligence is fundamentally different from human intelligence. Human intelligence requires motivation and a diligence both of which AI doesnt require as long as it has enough parameters in its model it can perfectly retain memory. This leads to a different kind of intelligence platform it makes sense in that perspective that it would be good in math because math requires knowing and retaining a large corpus of math to succeed and to scale and trying out different approach. All these things the machine can do easily.
>>543906581Sounds like some useless bullshit
>>543908694Hahaha fucking based retard>The Swarm: OpenAI deployed an unreleased internal model orchestrating a swarm of 10,000 concurrent AI agents working in parallel>The main search ran continuously for 88 hours (nearly 4 days), followed by an additional 17 hours of formalization and verification in the Lean programming language using GPT-6 Astra>The Output: The process generated 2.7 million messages and consumed roughly 130 billion output tokens for the Navier-Stokes subset alone (with the total multi-problem push exceeding 300 billion tokens>To understand the sheer scale of the Navier–Stokes run, this brief 4.5-day computing effort consumed roughly the same amount of electricity as:>1,500 to 3,000 average North American homes use in an entire year.>Running a large 150-megawatt data center at maximum capacity for the entire duration of the project.So, my bad, after doing some MATH, they used an entire LARGE citie's worth of electricity
>>543908222>>543908694
>>543908681No it isn't>Red things are redAckshully no, you haven't proven that yet>Proves it with 20 million dollars in computeWaow wow lookit this guise AI is insane - it's SUPERAI omgggMeanwhile, anybody not retarded just looks at red things and calls them red
>>543908974>Anonymous (ID: 2jlUDWmg) 10/07/26(Wed)07:21:12 No.543908974â–¶>File: AIshill.jpg (155 KB, 950x983)>>>543908222 (You)>>>543908694
>>543909258Computer scientists using Standard automated theorem proving (ATP) in the last few decades have already written highly optimized, non-AI C and C++ programs to exhaustively brute-force logical steps. The roadblock is always the combinatorial explosion; the search tree of math grows so fast that even the world’s fastest supercomputers exhaust compute after only a few steps. What the tech labs did wasn't a conceptual leap in intelligence; they just hooked up specialized "priors" to prune that tree, backed by millions of dollars in compute, to run the brute-forcing at a scale traditional C programs never had the compute to perform
>>543906540>and so on and so forth>*sHnrfff*
>>543908517>used 10,000 metric fucktons of compute to force a valid result>valid resultIn other words you're not disputing the solution just the efficiency of the process. No further discussion necessary.
>>543908728We don't need math
>>543909476retard
>>543906540but can ai explain it to retards like me?
>>5439065402 more twitter headlines and AI will pay back its investors
>>543906540>gives you the wrong answer
>>543906790theyre not even remotely intelligent, it a computer, it calculates. its not that deep
>>543909469>no further discussion need. Oh, fuck no buddy, we're unpacking this shit right infront of the kidsFrontierMath hosts open problems a pilot benchmark made of open research problems that professional mathematicians have actively tried and failed to solve. There is no existing human roadmap in Lean or LaTeX to distort the statistical weights, and the brute-force architecture hits a wall. Since its release, no AI has solved a single problem. 0%.A model's only skillset is executing a massive Monte Carlo Tree Search to find a known path, and when the path isn't known, AI eats shit into the dirt. It's a token predictor. The solution has to be already known in order for it to work.
It didn't though. All these accomplishments are lies. Like whenever china claims to have invented some new breakthrough. When you peel back the top layer you find out quickly how it's never true.
>>543907406I’m an unironic professional mathematician. I’m definitely not Tao and never will be but a lot of these results are directly interesting and relevant to me.
>>543909710They claim to put these proofs into Lean and have professional mathematicians verify them. How true that is I’m not sure, but I don’t think these results are made up.
>>543906581So humans just missed certain probabilities of logical formations?
>>543906581Ooooh it solved Dinky's 73/74 half pipe vertical theorum! Looks like I can quit my job now!
>>543909680This is closer to what actually happens than what most people think happens, these AIs are typically brute forcing approaches by doing the amount of searching it would take humans decades to do in a few days by burning millions of tokens. However, “ The solution has to be already known in order for it to work.” is just not true.
>>543908822>t-they spent 1 million dollars to solve a math problem! They spent 1 million dollars and 4 days, instead of 10 million dollars over 20 years!!!!!!
>>543909783Very stupid, inaccurate, but relevant explanation hereSay you lose a needle in a literal haystack. The mathematician approach would be:> Based on where the needle fell, and by taking into account how the haystack is shaped, the needle is very likely in this 0.1sqm space. Now let's use a metal detector and see if we find it or not.AI's "approach" is:> Let's check every single piece of material in the barn. Is it a needle? If no, ignore. Repeat until needle found.Does it work? Yes it does. Do you see how retarded that is? Can you call that "math"?Not saying this won't change in the future, but that's what AI does today. Same with software development - it just throws shit at the wall until it sticks. 99% of people involved in the process, including some software developers, can't really see that. So it looks like magic. The damage this is doing to all codebases in the world is incalculable. We will have to clean up this shit for decades.
>>543908822I don’t see how this is any cheaper then spending tens of millions or more in grant money for human mathematicians to walk down the same trial and error reasoning threads for decades. Those human mathematicians also use water and electricity and need their salaries and benefits paid for.I’m a mathematician but the cope around AI math is just getting sad and receding into ever tighter corners. Terence Tao is arguably the best among us and he understands that this is real and here to stay, so we’d better get to understanding our new role with this tool instead of bitching about it.
>>543909680>professional mathematicians have actively tried and failed to solve>no AI has solved a single problemYou're wrong on both points, not even doing a basic check on Wikipedia... so spin this however you like. Your anti-AI emotional freakout, is obvious.
>>543910060I mostly agree. Unfortunately however the single most expensive thing in developed countries is human labor, and no private company or academic journal has any incentive to watch out for future generations. So even with this sort of brute force approach that leaves shit stains on the walls, you can be sure that if it allows companies to hire less people and pay out less benefits, they’re going to keep doing it until the very moment the accumulated damage destroys them.
>>543910110>I don’t see how this is any cheaper then spending tens of millions or more in grant money for human mathematicians to walk down the same trial and error reasoning threads for decadesThat's not the argument. The argument is they're saying "AI solved" where AI did not actually solve anything. It's like marketing a flying car, but then saying it only flies while it's falling.Furthermore, you're not wrong - but theres a reason why we don't dump fucktons of grant money into this kind of research consistently - because there a better things to work on, like batteries or superconductors, but they aren't doing that because their models are not intelligent and cannot even perform math.>>543910201>basic check on wikipediaWhat did you just go edit that?The only thing AI is actually good at is porn and propaganda, anon
>>543906713I'm a janitor, and I use these equations in my daily work.
>>543910110> I don’t see how this is any cheaper then spending tens of millions or more in grant money for human mathematicians to walk down the same trial and error reasoning threads for decades. It's just just about "getting there". How you get there is mighty important, especially in math. It's also very "visible", in a way. The process really appears retarded if you take a close look at it.Same in software development btw. Just, the process doesn't appear as retarded there. Lots more people are buying into the hype, especially middle managers that didn't know better before AI and still know jack shit today.The amount of damage this is doing is incalculable.It's like digging a ditch by renting a backhoe vs hiring 100k jeets with a spoon. Is the ditch dug in both cases? Yes. > Terence Tao is arguably the best among us and he understands that this is real and here to stayAppeal to authority? Really?
I let a local nodel vibeslop me a script that it would have taken me six hours in about two minutes. Looks clean as fuck, commented immaculately. Not even mad about being replaced.
>>543910234Oh don't get me wrong here. This is 100% the end of white collar work.I'm just saying that people that believe journalists (imagine that) are made to think that "AI solved math problem X", but it means something different than how they make it sound like.Doesn't mean that this doesn't have an economical impact. AI doesn't need to be genius level, or even human level, to have economic viability - it just has to cost less than equivalent humans and do the task up to an acceptable bar. This is already happening, today.People are looking at what AI does in math and compsci, but those are the wrong place to start with. They're possibly the hardest field of application. There's still a need for novelty there, and plenty of it. AI struggles with that.Look somewhere else - engineers, architects, accountants, lawyers, middle management, doctors, functionally all replaceable TODAY with the SAME or BETTER level of accuracy for 99% of the cases. Most of those professions don't really "invent" shit, they just apply some kind of pattern to a known problem. AI can do that, and much better, for much cheaper.I mean, even just 10 years ago you had to PAY for translation work. That is dead and buried. Everything else in the above list will go the same way. Being an architect will just be economically nonviable.
>>543909510Yes, that's the only thing AI can do.Yes, you are extremely retard.
>>543910436> I let a local nodel vibeslop me a script that it would have taken me six hours in about two minutesIf it would have taken you 6 hrs, it means either of these:> you're a really bad developer> script is complexchances are you're just a bad developer. If you're good, and you take a look at the script, you'll see that's an unmaintenable ball of shit. It works, sure. It's nothing like it would have been if someone competent would have written it.Now imagine that the whole world is shitting out that, millions times a second, and embedding that in very critical software everyone uses.
>>543906540
>>543908517>>543908164this retard thinks you can bruteforce these problems like a fucking password cracker tool does lol what a terminal retard
>>543910853it is kind of the same approach memeflag kikeI wonder why you glowniggers are trying to push this narrative
>>543910900do you even know absolutely basic shit like what a Lie algebra is
>>543910919sure, now can you remove the memeflag?
>>543910769do you think a barbarian with an axe can defeat a murder drone
>>543910955sure thanks for telling me you're a retarded wop
>>543910978> eglin faggotimagine my surprise
>>543907375>Checks flag>Local plummerAre trying to solve India's toilet problem?
>>543910546So we agree. Retarded journalist and OpenAI PR departments are misrepresenting the nature of these discoveries, but that doesn’t change the fact that the writing is on the wall and we have to adjust to an economy dominated by AI displacing a lot of human labor. I don’t think pointing out Terence Tao’s attitude is an appeal-to-authority fallacy, it’s an appeal to competence. He understands the field very well and his assessment is that the way we do mathematics is going to be changed by the easy availability of a machine that can generate examples or test hypotheses en masse. AI still requires a LOT of guidance from a human expert to do research with, especially if you’re not able to burn trillions of tokens. This is also true for software, but the difference between math and SWE is that a result that “looks about right” and cuts corners to get there in math is manifestly wrong immediately , while in software you might be able to get away with it for months or years before anyone notices.
>>543911229Yes we doI'm just right now skeptical when it comes to AI's potential to do real innovation. What we're seeing today is infinite monkeys with a typewriter.
>>543910853The hypergraph incident (as another ill informed retard drew attention to) became a case study in exactly what i'm describing: an AI lab taking a brute-forced solution to a highly niche, low-notability problem and framing it as a historic leap toward AGI, only for the actual mathematical community to quietly scale back the achievement - it just found an algebraic construction that humans found too tedious to compute manually
>>543911435navier stokes was brute forced? lol how stupid can you be. again do you even know absolute basic shit like what a hilbert space is? and you're speculating that all of this was the same as when you used a tripcode to generator to tripfag on /lgbt/
>>543906540>>543906581so mathstards were being paid millions to solve utter drivel? whip em through the streets
>>543906723OPENAI won't release any result that could accidentally, directly or indirectly tarnish the legacy of any Jewish scientist
what difference does it make. the universe is 14 billion years old. no sign of superintelligence anywhere. if those people who think AI leads to an information singularity or w/e it's called then why don't we see one? probably because technology can only get you so far. at the end of the day you are still bound by the laws of nature. that means we can't get around in the universe, because of the huge distances. we are destined to rot and die in this solar system when the sun eventually exhausts its fuel.
>>543911498sit down you fucking tard, adults are speaking
>>543911616you don't know what an octonion is and call others tards lol the peak of Dunning-Krueger
>>543907509That would be classified.
>>543906581Translation 4 Plebs:Riemann Zeta can be shorthanded.Landau-Siegel:Proof that while Godel is right about unknown unknowns, all unknowns can be approximated.Unique Games:Approximations can be made at pace. The "Vibe Coding" of the Mathematical world. "How do these numbers makw you feel?"Mathematical W:Matrices will be more complex yet stable to correct with base approximations.Hodge CM:>>>/cm/Thompson Group F:"All failures outside our temple walls are unspeakably Evil."
>>543909401LMFAOIt’s hilarious not only that you used Grok to back up your anti-AI stance, but that Grok's response literally refutes your argument.The quote you shared says that:>What the tech labs did wasn't a conceptual leap in intelligence; they just hooked up specialized "priors" to prune that treeIf you're conceding that state-of-the-art models use built-in intelligence to stop combinatorial explosions, your argument falls apart. Narrowing down massive pools of possibilities is the exact definition of intelligence, and that's exactly how these AI proof methods work.
>>543907852Yes and they consider themselves the hardest to fool lol
>>543911535I started noticing that every scientist worshiped for his "brilliance" was a jew and I couldn't stop noticing.
>>543907601>So do any of these proofs gave actual practical applications or is this just more theoretical fart huffing that's the mathematical equivalent of moral philosophy?Yes, it fart huffing like Rationalist philosophy and quantum physics
>>543906540So where's the joy of discovery and learning if AI is just gonna do it all? Sounds like this shit will be the death of the pursuit of knowledge
>>543912022He said "priors," not spooky "intelligence." He could have equally said curve fitting based on training data.
>>543912197the existence of ikea doesnt stop you from making your own furniture
>>543911498Yes. Terrance Tao already outlined an approach even before this happened. It might not be just brute forcing calculations but it is brute forcing certain class of solutions
>>543912211Priors and intelligence are equivalent, numbnuts. The reason why you don't have to brute force your way through a dictionary to parse this very message is because of your prior knowledge of Inglés, i.e. a kind of intelligence. For similar reasons mathematicians are able to do their work: their priors inform their mathematical reasoning, i.e. their math intelligence.And even though this relation between priors and intelligence is nuanced, the leaf's/jeet's attempt at trivialising the results as mere brute-forcing is refuted in any case.
>>543906713You've never mixed baby formula portions before, have you, son?
>>543912422what he said could apply to some problems but not all 400+. and didn't apply to NS. you have to understand the wop and leaf from earlier are really stupid and actually think it's doing guess and check on possible answer #1, #2, etc. they actually think its some trivial thing we couldve done 20 years ago with the same amount of compute and nobody thought to do so because cope reasons
>>543906713I am a farmer and this will help me bigly
>>543912022>It’s hilarious not only that you used GrokGemini, actually. I can google things, you knowYou must be that really annoying janny that prunes threads every time you lose an argument, i can smell it>If you're Yeah see, you don't even understand what i'm asserting,>Narrowing down massive pools of possibilities is the exact definition of intelligence, and that's exactly how these AI proof methods workThis is actually just common sense, and the limiting factor here was never intelligence, it was available compute
Can AI divide by zero yet?
