The singularity is hereChatGPT 5.4 just solved a 64 year old math problem no human could ever solver before
>>108701674How do you know the answer is correct if no human can solve it?
>>108701697You plug it back in
>>108701697whenever a real person produces a novel mathematical proof or solution, that person is, for a while, the only person who knows the solutionof course other people can then step in and verifythe same applies to AI
>>108701674>solved shitty old math problemwho da fuck cares? just gimme cheap RAM!
>>108701674erdos problems are little problems for starting students. They've likely been solved many times by grad students who never took the time to publish.
>>108701674absolute state of math education of mutts if they fall for this shit.
>>108701828Some are simple, some are extremely difficult even for the most big brain mathematicians and still unsolved.
>>108701860and this is an extremely simple one. A few of them that are collatz inspired. Until an AI can literally solve a millennium problem or create new mathematics it's useless for math. Since the only point of doing these problems is to create new math that might be useful.
>>108701674Use case?
>>108701674>math problemsboomer cringe, not a single person with a life gives a shit
It didn't solve shot. It vacuumed up somebody's method from social media and spit it back out. It probably spit out 100000 incorrect solutions that the ai shill masters neglected to report on, and for the press involved when it incidentally and finally token puked out the correct one.
>>108701902Being unsolved means it wasn't a simple one.here is the solution:https://chatgpt.com/share/69dd1c83-b164-8385-bf2e-8533e9baba9c
>>108701946Unsolved often times means it's not worth solving. >ai, add infinities until you have infinite infinities
>>108701932sybau dunning kruger mong
>>108701932ESL brainlet kekLrn2 past tense
>>108701946There is no new math in that solution. Ergo it's uninteresting.
>>108701971>>108701980The problem was deemed interesting by superduper mathematician paul erdos himself. I trust his judgement over yours.
>>108701976>esl Only browns and faggots have s○ygasms for ai. A white chatbot cheerleader is literally unheard of, unless you are part of the grift. Have fun drinking from Ganges.
The LLM cultists have said this so many times, and every time, it ends up being a lie.
>>108701697Jesus you didn't get past calc 2 did you
>>108701971>>108701674>>108701828This is bullshit to begin with. You people do realize this is how quantum physicists solve "problems"Most of the time is just "hey, let's just add another dimension to the equation" ... nigger really ?, ARE YOU FOR REAL ?! Ye fucking brainlet....At this point, quantum physics is more esoteric than astrology.One shutters to imagine what horrors manifest if Zoomers touch math problems on grand scale.
>>108701992>TRUST THE SCIENCEEuclid was blown the fuck away by squares drawn in dirt. At least his discoveries related to it advanced understanding instead of, for this problem in particular, being a brute force computational waste of time.
>>108701674cool. is the math settled on R's yet?
>>108701697these are the people saying AI is bad and that it's taking up all the heckin water and we should stop building those darn datacenters
>>108701971please be trolling. for my own sanity i can't believe people this retarded actually exist.
>>108702029The vast majority of maths done since the 60s is completely useless
>>108701992Interesting enough to explain to a child/highschool/graduate student and walk them through a solution attempt. Go read about Erdos life and work. I scrolled through the entire thought process and it's actual intelligence is still just as low as any other 1~10T parameter model. If there simply didn't happen to exist, a straight forward solution, that just required more modern tools from probability theory. Erdos wrote that problem in 1935 and some of the arguments popularized in the 2020s
>>108702071>If there simply didn't happen to exist, a straight forward solution, that just required more modern tools from probability theory (it would have likely not solved it)
>>108702071Holy ESL wordsalad, get the fuck back to twitter ranjeesh
>>108702071good morning sir
>>108701674>retards who never gave a shit about mathematics before come crawling out of the woodwork to praise their lord and savior, the great equalizer, the machine that makes their miserable lives less miserable because it ruins everyone else'sit's ok. you don't know better. it's not your fault
>>108702078>>108702081Holy not an argument
>>108701712Anal plug?
>>108701946No, there are an infinite number of problems that have not been solved yet. This is how phd students manage to generate papers most of the time.There are also an infinite number of problems that are unsolvable in any consistent system.
>>108702102i've already redeemed it, there's nothing you can do
>>108702134But there is only finite number of ERDOS problems. The clout you get for solving those is infinite.
>>108702150LOL no. Please stop talking about shit you don't understand.
>>108702217literally nothing he said was wrong
>>108701674>>108701745>>108701946>>108702035Wrong answer, another made up mathematical formula
>>108702217Are you arguing that there Erdos wrote an infinite amount of problems during his lifetime? Fucking retard
>>108701674>it took 80 minutes and consumed 10 rainforests to conclude 1/(a log(a)) is less than or equal to 1/(a log(a)-c)
>>108701674just proves that there's zero intellect involved in math or chess, just pattern matching
>math that human can't solveliterally impossiblewe all know all math problems are made up by human themselves to credit themselves for the history books
>>108701946>he believed itLLMAO
>>108701926t. gas station employee with a "life"
>>108701674How is this being solved going to make my life better?
>>108702720retardmaxxing
>>108701674give it time and someone will find an obscure piece of literature that contained this solution exactly
>>108703155Oh yes, it will turn up.Probably not translated from russian, chinese or hindi yet. Quite possibly proved hundreds of years ago.
>>108704071>russian, chinese or hindi>invention>less than 2000 years oldpick two
>>108703155it's called the library of babel
>>108702720some math nerd feels bad because he will never get credit or citations (the most important part)
I wonder how many unsolved math problems exist because they're not interesting or useful so no one bothers with them.
>>108701674>ChatGPT spammed WolframAlpha for 80 minutes until it got a "solution">the "solution" being an abbreviation of the first lineI should thank god that I am not simple-minded enough to be as easily impressed as an AI jeet. These are the kinds of people who 500 years ago would've tortured and burned a woman as a witch for owning a black cat or some dumb shit like that.
>>108701674>unc shocked by clanker vibe-coding mathslop
>>108701674gemini got mogged kek
>>108701795>computers finally computes something interesting instead of adding digits to pi and drawing triangles.>people cries about itAI can just do all the necessary thinking it needs to make hardware even cheaper and make itself easier to run. It will take some time but in the end hardware is going to get better and cheaper at a higher pace than before the shortages.
>>108702425>watches nasa livestreams
>>108704927yes, seethe
>>108701697lol
>>108701674Crazy that somebody is willing to pay 100 bucks to solve a 64 year old math problem.
>>108702021The Copenhagen Interpretation and its consequences have been a disaster for the human race.
>>108702021why can't you people just say you don't understand and shut the hell up
>>108701697>this bsod's the npcs
>>108701946>I’m polishing the prime-power Mertens lemmaa sentence dreamed up the utterly demented
>>108701674Asking an LLM to do math is like asking a deaf retard to write a symphony.
>>108701674The scariest thing about AI solving math problems is that humans are hardstuck by their cognitive ability and there's some invisible universal numbers we havent discovered yet that makes this possible for AI to solve but not individual humans.
>>108706818AIs can't solve math problems. They aren't calculators, they're prediction models.
>>108706829>It's not X, it's Y.
If you have solved fewer erdos problems than chatgpt, you aren't allowed to criticise it. Them's the rules.
If you waste your time solving math equations which don't have an application then why bother.There are currently unsolved math problems which have entire industry changing potential already realised and accounted for, all someone has to do is solve it and get it to apply correctly and boom, we now have AV2 or whatever.Wasting ai to prove a point is stupid, use it to make money, lots of money
>>108701697Ask it to write a proof in Lean>>108706774Cope
>>108702100>mathematicians NEED to be able to work on these retarded, useless toy problems for their entire 80 year lifespan, just throwing government money down the drain by the billionsFuck mathematicians, fuck physicists, fuck scientists of all kinds. Nothing but scammers everywhere. Imagine having infinite fucking government grant money and still not having working fusion power.
>>108707132today's useless math may turn out to be unreasonably effective at describing newly observed and predicting yet to be observed physical phenomenons tomorrow
>>108702021Quantum relativity is where it's at.
>>108707865how much % of math is totally useless as opposed to later "discovered" to be key solution to some other shit, making it useful
>>108706911You do it for the love of the game
>>108701674probably found the answer in some 2006 post from an obscure Uzbek math phpBB forum.
>>108708468I mean specifically using AI on them
>>108707865>"may"you "may" pay me to jerk off and eat burgers for 80 years, it "may" reveal something about our universe that was previously undiscovered and also uselesswhile I'm at it, who gives a fuck how stars move or what constitutes their current atmosphere. if I were paid to look at shit in the sky more that a million miles away I'd be deeply embarrassed and depressed. a life wasted for absolutely no return. everyday spent producing research and paperwork with no use, no purpose.anyone that engages in star gazing is paid to waste their intelligence looking at useless shit so they don't turn their big brains to something actually useful for humanity
>>108708540>UzbekNo, it's probably some indian middle schooler's homework
>>108708665you're going to spend your miserable life jerking off alone and contributing nothing of value to society, thoughbeit
>>108708702you seem upset, do you get paid to look upLMAO
>>108706713$100 for $100,000 of advertisingProbably broke even with all the attempts, but it's still a good headline.
Read the article until the end brownoid. It didn't actually solve it and actual mathematicians had to rewrite almost the whole thing. The only thing it did was give them a rough outline for the approach that you can use to solve it.
