OpenAI is gearing up to attack the Riemann Hypothesis and P = NP.Navier-Stokes took around 10,000 Astra agents over the course of 88 hours, or about 10% of the company's total swarm. Most of the swarm is working on developing the next model, but they seem to have slowed development out of fear and are devoting larger fractions of the swarm to STEM problems. Do you think these problems will be toppled with tens of thousands of agents, or are they too hard?
>>17051662Another vibeGOD victory
>>17051662>Do you think these problems will be toppled with tens of thousands of agents, or are they too hard?AI conquered art. Which sci-fi movies always claimed was the hardest job for machines.We can deduce that Math is now piss easy in comparison.
I'm not ready to live in a world where N = NP is true. What would that even look like?
>>17051662With Riemann I think they would have to give a proof that all of the non-trivial zeroes are one the critical line, so brute forcing a counter example wouldn't work. But maybe showing that a counter example doesn't exist through brute force is the way. So then I don't think it should be that hard.
>>17051699If P=NP, then under certain conditions, civilization just ends. Cybersecurity collapses at the same time that AI recursive self-improvement gains an ENORMOUS capabilities multiplier, as it can search the design space vastly more easily.
sorry but neither of these can be brute forced like n-s counter examplethis wont produce a resultt. pro
Oh my God who fucking cares?
Even if P=NP (which it won't be), we still need to find an algorithm to solve an NP-complete problem in polynomial time. That algorithm could have factors of 10^10^100 in it making it impractical for real use.
>>17051701Yes, this is what I meant by 'certain conditions'. >>17051703In surveys, most mathematicians think P=NP has like a 5-10% chance of being true. That plus a generally useful algorithm existing: Maybe 2%. So most likely we're fine in that one narrow case, but I strongly doubt superintelligence is survivable for other reasons.
>>17051710Meant this post for >>17051706
>>17051705I love the salty luddites.They trash talked AI for years, now they can't handle when it fights back. Literally David vs Goliath.
>>17051662They aren't solving anything because Navier Stokes only worked because two random Spaniards (Diego Córdoba and Martínez-Zoroa) had a scalable superoriginal approach, and they don't have anything like this now for these two other problems, which will show that Córdoba and Martínez were the key behind OpenAI "proof".
>>17051662Why does it always has to be math?Please unify general relativity and quantum mechanics, explain why there is more matter than anitmatter, do cold fusion or invent some other clean, cheap renewable source of energy, invent mobile cells that lasts a week etc.Why pizza is better than AI? Pizza can feed a family of four.
>>17051717>Physicschads filtering tokensloppers once againBased
>>17051715Someone screenshot this comment.Don't let them lie and twist history. Whenever AI makes an achievement, they're going to pretend they were never on the wrong side of history.
>>17051662we need to pull the plug now immediately before it finishes mathematics
>>17051722The Genie is out of the bottle.You can't stop it.
It's joever. This board can be deleted soon.
>>17051727Calling it now: P=NP will be solved by Christmas 2026.I want to see the meltdowns.
>>170517152 millenium problems will be solved by the end of the year luddite. What will be your excuse then
>>17051722>finishesif i loved math i would be thrilled to learn all the cool shit it will output and if I were smart I would be thrilled about how far mathematical theory can be pushed. Only luddites that do math for ego are hating this.
>>17051662What if they used their super computer to like figure out how to regrow limbs instead of stupid faggot nerd puzzles?
>>17051748AI is being used to design boutique drugs that cure rare diseases. Enjoy the ride you atupid fucking luddite, CHATGPT might cure my mother's multiple sclerosis and you only care about your fragile ego.
>P=NP>P/P=NP/P>N=1where Nobel?
>>17051751I do not care about the "prestige" associated with solving stupid math problems. I care about using AI systems to solve practical problems like regrowing limbs or as you mentioned drugs to cure disease, you illiterate faggot.
>>17051752There is no Nobel in math because he hated it.
>>17051744>by the end of the yearNah, end of the month at most. Screenshot this.
>>17051758>because he hated it.kek so he was a brainlet
>>17051764Maybe he just intuitively knew it wasn't worthwhile. I don't think he knew AI would end up solving it first but some people are good at acting on hunches like this.
From now until the end of time we get given infinite universal truths on tap yet no one knows why they are true nor can anyone ever know since each truth is built upon 10,000+ years worth of ten trillion AI agents' work. >ah yes but we humans get to decide how to use these truthsDo we though? The same black box godlike intelligence that produces the universal truths for us can also be asked how best to apply them. For any human suggestion, would society not be better served by leveraging the AI to implement that suggestion in such a manner that maximises the intention while minimising human suffering?Where do we, humans, actually go from here?
>>17051763probably you are right kek
>>17051753but it is already doing that. Clause designed a drug that can cure Alzeimer's.
>>17051799No they didn’t
>>17051662>The smartest mathematicians are tackling N-S>They upload their work to chatGPT>openAI steals their work>The smartest mathematicians are surprised by thisopenAI will solve these problems only after genius mathematicians solve these problems. """"AI"""" is retarded, and will never create anything new. Gayass faggot clanker-lovers ITT need to KILL THEMSELVES. Lol. Lmao haha. hee hee I'm feeling whimsical fellas.
>>17051717>Please unify general relativity and quantum mechanics...You assume this (like curing cancer) is a scientific problem. It is not. It is a political problem.
>>17051822You may discuss it on /pol/.I guess,
>>17051803I don't get it though, if AI is doing the work either way, why do we care? I mean if anything, it seems more honest for openAI to take the credit.
>>17051802you are behind the times unc>>17051803When it solves Cobaltz conjecture what will your cope be
N=NP>N=NPN=NP>N=NPN=NP>N=NPN=NP>N=NPN=NP>N=NP
>>17051662>Situation detected Is this the new way to say happening without getting filtered?
>>17051662AIChads I kneel
>>17051717>Please unify general relativity and quantum mechanicsThey are going to try and do this and accidently disprove relativity instead. Just you watch.
>>17051662>P=NP Dont do it mathbros, that would be pretty bad for my bitcorns
> P != NPThere, I've solved it.
>>17051951>and accidently disprove relativity insteadhad to check what board i'm on
>P=NP>N=1there, my one million dollar please
>>17051959Do anons on other boards even know about relativity vs aether autism? Figured its a /sci/ special.
>>17051662>P = NPeither N = 1 and/or P = 0where's my nobel prize?
>>17052178sorry anon but I was the first to solve it, you just copied it from me >>17051752
>>17052182you merely assumed N to be 1 but you never said that P could be 0 so my solution is more complete
>>17051838>High schooler uses my casio calculator to solve an equation>May as well credit casio with every high school diploma issued after the year 2000Lmao @ the level of cope here>>17051839>Cobaltz conjecturePretending to be retarded? AI hallucination? Hmmm.(At best it could do a disproof by counter example).
If they basically stole NS will the model be able to do these other ones? We will see....
You guys are misunderstanding P=NP. They're not trying to prove P=NP which would be absurd and game changing, they're trying to prove P!=NP which is what everyone already believes to be true and wouldn't change anything other than settle the problem. No one in computing actually thinks P=NP
>>17051703even if P=NP, it could be the case that an algorithm connectin the two only becomes useful with prohibitively large input sizesi'm talking about not enough atoms in the universe to encode such a number sizes
i come from a mathematical universe where we we cryptopgraphy is impossibleyou guys have no idea what's in store
