So who's right/g/? Can chess be fully solved with enough computing power?: https://x.com/elonmusk/status/2095868436597354751
>>109733466Bob would have agreed with Elon.
>zomgf there are too many position!Retarded take. Elon is right.
>>109733466"Fully solved" no, if we're being precise about what fully solving means, Technically unless you iterate every move it's not possible to know a bad line of moves won't magically turn optimal.Effectively solved - yes it pretty much already is with the best AIs typically drawing every game
>>109733500>Technically unless you iterate every move it's not possible to know a bad line of moves won't magically turn optimal.thats literally means enough computing power would solve it
>>109733466I thought they already solved chess with Deep Blue.
>>109733466>So who's right/g/?one literally uses the handle CHESS.com, the other is coping with — *checking the fucking notes* — skill issue since childhoodget therapy, elon
>>109733466>many games/sportsSuch as >uhm checkerslolLMAO FAGGOT
>>109733510That goes for every problem. But we don't have infinite computing resources and energy in the reachable universe for it.
>>109733490>>109733490>samefagging your own thread to suck on the cock of a loser that was begging for an invite to epstein's island like it was a private torrent tracker and got BTFOlol>Bob would have agreed with Elon.he hated people like elon lmao
>>109733500>it's not possible to know a bad line of moves won't magically turn optimalThat seems like statistical bullshit along the lines of "every shuffled deck of cards is unique", where it's less about whether bad moves can be salvaged and more about fucking about with the definition of a bad move until the desired hypothesis is achieved.
>>109733466How about Go?
>>109733496Elon is never right.
>>109733611>>109733500>replying to yourself with ai slop just so you can worship a loser pedophile that everyone laughs at it's amazing how cringe losers like this exist
>>109733516I think they solved Go but not chess.
Standard goychess is for brainlets. No skill needed except for memorization. Real Aryan's chess is 960/Fischer Variation.
>>109733598compounding superposition states could get us there>inb4 sci-fiye, ye ik. fuck off, let me dream
>>109733627Stop trying so hard, you're gonna blow a fucking gasket.
>>109733466Chess is fully solvable you just need a lot of compute and storage.https://www.youtube.com/watch?v=fJMXkBpN4MQ&t=242
>>109733634Everyone who plays chess but doesn't enjoy strategy games is a poser
>>109733466Why should we care about his take on [anything]?
>>109733466Standard 8×8 chess with 75 move rule and fivefold repetition can be optimally last upto 17697 moves. Generalised n×n version is pspace complete but that doesn't mean a single instance might be intractable. It might be possible there's a small description for the optimal strategy.
Oh no no no...
>>109733466Of course. It's already nearly impossible for even the best humans to get even near mediocre chess programs.Chess has not that many moves possible so you can feasibly store everything with some pretty big disks and do a search on the moves. Right now we have pretty huge databases for pretty deep early and late moves, it's only the middle we use algorithms for.
>>109733466Elon is right, trivially so, and everyone agreed with that statement as of 1d ago, OP's local time. The only reason anyone is disagreeing now is because le rocket man said it, so their sense of self identity requires them to say "nuh uh!"
>>109733598Some problems are in continuous space, some problems have space states so large even every atom in the universe combined wouldn't provide enough compute or capacity to solve them exactly by these methods.
>>10973384420 years ago the exact same statement musk is making today was widespread and accepted by the chess computing community.
>>109733617Can someone post any one of elon's public predictions that was true?Since musk is a 180iq billionaire tech genius this should be trivial.
>>109733622Go is solved less than chess is.
>>109733633Yeah, I could see it being technically possible through quantum computing, but then it comes down to actually finding the right equations which is probably extremely, extremely complex. And of course, quantum computing is very difficult for us to do.
>>109733804No Elon bros nooooo
The usual wisdom is that the numerical possibilities for moves go into such high numbers its difficult to calculate for a normal computer. You have to limit the calculations at some point to get a result and that means its not """solved"""But llms have so many parameters its not specially impossible task I guess.What a chess account employee or elon musk have to do with that specific argument Ive no idea. American left right politics probably
>>109733946>llms have so many parameters its not specially impossible task I guess./g/ - tech illiterates ass post
>>109733466He's correct. Chess is a finite problem so with enough memory and logic the entire chess moves can be known and solved.
>>109733959what other thing could it be aboutregular pcs? Who gives a shitLLms? Those operate on their parameters. If you do something interesting with chess you have an llm with a lot of chess parameters.
>>109733466Is that your gay lovers twitter you posting here?
>>109733466It has already been heuristically solved anyway. Even alphago btfo'd real pros in Go and there are even more "valid non stupid moves" there. So idk wtf these cards are on about.
>>109733610Hate and disagreement are two different things.
>>109733946Please say you are just pretending to be retarded.
>>109734055I already replied to you above, please move on.
