ChatGPT Broke Chess

➡️ 25% OFF ALL COURSES: bit.ly/tacticscourse
➡️ Ice cream signed book event: bit.ly/icecreamchess
➡️ Get my best-selling chess book: geni.us/gothamchess
➡️ My book in the UK and Europe: bit.ly/3qFqSf7
➡️ Mein Buch auf Deutsch: bit.ly/45fKt3R
➡️ Mi libro en Español: bit.ly/3Y5xaRx
➡️ Start Playing Chess FOR FREE: bit.ly/3Xa3EsB
➡️ Enjoy my videos? Donate Here : www.paypal.me/gothamchess
Email me your games: [e-mail]
Sponsors, Business, Media: [e-mail] - [DO NOT SEND GAMES HERE]
⭐️ Follow Me If You Are Amazing:
➡️ CAMEO: www.cameo.com/gothamchess
➡️ SNAP: / levy.rozman
➡️ INSTAGRAM: / gothamchess
➡️ TWITCH: / gothamchess
➡️ TIKTOK: / levyrozman
➡️ TWITTER: / gothamchess
➡️ GOTHAM DISCORD: / discord
➡️ THUMBNAILS BY: / jchessnoob

[@elenalikerapmusic]

chat gpt chess content is so dumb bro

[@bryansmith7238]

PIN OF SHAME

[@lightningmck5158]

PIN OF SHAME!!!

[@mymsoup]

shin of pame

[@GamezK734]

PIN OF SHAME !!!

[@elenalikerapmusic]

i deserved this dammit

[@NimbusCloud_]

ChatGPT 3.5: plays *illegal moves* accidentally ChatGPT 4: plays *illegal moves* on purpose and *makes your moves for you*

[@theBrunoLYNX]

Fr

[@NathanielMahRuiYangAcsj]

Lol 😂😂😂

[@samiraperi467]

ChadGPT

[@JJKGOJOSHOT]

Which ones scarier? 😂

[@Evex6]

Levy: Find's a forced mate in 5 ChatGPT flipping the board around: 𝗜 𝗮𝗺 𝗻𝗼𝘁 𝗶𝗻 𝗱𝗮𝗻𝗴𝗲𝗿 𝗟𝗲𝘃𝘆, 𝗜 𝗔𝗠 𝘁𝗵𝗲 𝗗𝗮𝗻𝗴𝗲𝗿

[@someverycool4552]

Show me a stockfish who can do that

[@Varscool]

ChatGPT : You are under my Genjutsu

[@firstlegend5105]

Slide 2

[@obsidianquill7949]

@@Varscoolchatgpt really flipped the script like itachi flipping kurenais genjutsu 💀

[@Varscool]

@@obsidianquill7949 😂

[@stealthbeastgaming]

19:07 - "And then Chat GPT decided to sacrifice... THE RRRRRRRRULES"

[@nicholasleclerc1583]

Lmfaooooo !!!!!!

[@leventeiles2281]

Underrated comment!

[@LiteraIIy_Nobody]

And then Chat GPT decided to sacrifice...MY BRRRRRRAINCELLS!

[@andrewkosenko2757]

19:20 Props to chatgpt for finding the best move and going diagonally with a rook to c4

[@CanakoMC]

ChatGPT is definitely one of the chess players of all time.

[@MichaelZankel]

It’s not, it’s an AI. Not an actual player.

[@biz6361]

@@MichaelZankelwhy u so uptight bruh

[@melkzaih3437]

​@@biz6361ong

[@brentgroen3204]

​@@MichaelZankelyou must get all the babes and male babes like that

[@mitchellgardner9201]

its still a player.@@MichaelZankel

[@Alex-cw3rz]

Moving your king away from the pawn promotion was hilarious, It's as brazen as someone in in person chess going "what's that over there" to their opponent and then moving the king while their opponent is looking the other way

[@PaulDormody]

AI has learned to cheat.

[@thurgasivaruban2511]

😂😂😂

[@123crackera]

​​@@PaulDormodyyeah... But does it know how to rage quit?

[@youtubeforcinghandlessucks]

@@123crackera I’m sorry but I prefer not to continue this conversation. I’m still learning so I appreciate your understanding and patience.🙏

[@kamikeserpentail3778]

It was like, last turn I gave you a free move, this turn I'm taking it back.

[@zerorusher]

11:40 its very likely the chat had reached its context window limit. It means its "memory is full" and it has to summarize parts of the char in order to continue, hence the "memory lapses".

[@YuraSuper2048]

that's stupid. gpt-4 has 8k tokens of context memory that WAY enough to store a conversation this small

[@emphyriohazzl1510]

No it s not that, and it still doesn't undertsand the rules at all (despite being able to give them in all details if asked). It will make these illegal moves even if you keep telling him where every piece is after each move. It's only able to play rather well in thz opening in main lines because it has access to a lot of text sources mentionning common opzning lines. It just plays by mimetism.

[@ChaseAnderson5k]

He did tell it the moves again so people are correct. Though, people are probably forgetting if it can play nearly 30 moves correctly, then assuming it's good at more than just this opening there's a good chance it could defeat many of the people saying it's bad without breaking rules. But it also may not know what to do when people make bad moves, so who knows what would happen, it may start getting confused faster in a 1000 level game.

[@YuraSuper2048]

the model just got confused, either by random sampler or by itself, that's all there is to it. gpt 4 is a very big model but it will still make stupid mistakes.

[@Bixmy]

​@@YuraSuper20488k token is not as much as you think.

[@Quartapple]

This is legitimately one of your best videos. By the 21 minute mark, I was ENTHRALLED. What a fantastic saga that match ended up being

[@Barcamuda]

Chess speaks for itself

[@AziDoesQuestionableThings]

Nans Hiemann

[@Hudoi-1]

Nans Hiemann

[@lucastheg6156]

Nans hiemann

[@Lucaamroleitner]

Nans Hiemann

[@jacobnunya808]

Nans Hiemann

[@ArguedLlama]

I’m glad Levy finally incorporated chess into his channel!

[@vikramadityakodavalla3795]

what the

[@DPME820]

failed attempt for the pin of shame

[@ArguedLlama]

@@DPME820 how is this an attempt at pin of shame?! 😂

[@ammarahmed7170]

Me too!I really wanted Levy to start making chess videos.

[@trapezoided]

​@@vikramadityakodavalla3795 how is it even remotely close to being an attempt for a pin of shame?

[@svendaman925]

honestly if chatgpt gets really good it could be such a great tool for analyzing games as its able to explain intentions in words, which is not something modern chess engines can do

[@peymanafshani9304]

chatGPT can't do anything like that. It's a chat bot best suited for making word salad.

[@TheBlueArcher]

@@peymanafshani9304 I'm not so sure about that. Train it on enough high level games and analysis. the word salad will be predicting the words that describe a moves intentions pretty accurately. I imagine it would be a more human-like player than an engine as well since it would presumably be likely be trained on human games more than engine games. Even this video, yes chatgpt has memory, but i'm surprised a language model can actually somehow end up predicting so many legal moves for so long without actually storing a chess state in memory is petty amazing in and of itself.

[@GurgleGurgleBlahBlah]

@@TheBlueArcher Chess is primarily a combinatorial game where the quality of a move in most cases depends on developments many plies ahead. A language model can do very well in openings where there is a lot of human games, analysis and text available for them to produce text but as soon as you get to midgame it basically will be useless and very generic, as shown in this video. chatGPT is basically a smart search engine that spits out text based on text that is has been fed to (i.e., known information). It can't produce deep novelty, it can't think and it can't reason and it can't explain intentions because the model has nothing inside it that is capable of those things. It cannot analyze a novel position the same way stockfish can.

[@Nomad43944]

​​@@GurgleGurgleBlahBlahthere's no AI that 'thinks' and reasons' true AI dosent exit atleast not now sophisticated codings to perform specific tasks is not intelligence

[@ronaldking1054]

@@GurgleGurgleBlahBlah It reasoned well enough to determine which piece was critical enough to kill even through an unethical, illegal move.

[@hjtomi_]

The way you tell a story about a chess game in such an interesting manner is still mesmerizing for me

[@jeffbarely5168]

ChatGPT never fails to break chess

[@sirstamfordraffles6557]

_What do you mean?-It played flawlessly!_

[@dontbanmerage416]

I see. Very impressive that levy can do this while recovering from being absolutely rocked

[@saheranwar64]

wdym?

[@georgetown2180]

@@saheranwar64 rock you like a hurricane

[@logos706]

What do you mean?

[@TroyLan]

"I will not resort to tactics used by ChatGPT" 😂😂😂😂😂😂😂😂

[@vukasingames5772]

Chat gpt 4 when it starts losing:Chat gpt is back

[@akiroclimbs2300]

This was insane. I remember watching those old ChatGPT videos over the summer seeing in manifest pieces from the fifth dimension and fabricating it’s own rules. I never plan to purchase ChatGPT 4, but the jump from the previous generation chat bot to a solid 2200 chess player is actually scary.

[@jacobnunya808]

It is not a 2200. It played a lot of good moves but the blunders would sink its rating well below 2000.

[@akiroclimbs2300]

@@jacobnunya808 when i made this comment i hadn’t finished the whole video when it went bonkers

[@keypey8256]

GPT 3.5 instruct which is a version which wasn't trained to chat apperently is on 1800 elo level when given current position on each move using PGN

[@ishaanm8270]

Bing AI uses GPT-4 so you may be able to test it thete

[@keypey8256]

@@ishaanm8270 yes but due to fine tuning and prompt engineering, Bing has a tendency to perform worse than the actual gpt 4

[@lio1234234]

You should totally use the vision version of GPT-4 and upload an image of the board from the point of view of its position for each move! I'd love to see how that turns out, because then it isn't even relying on chess notation anymore!

[@greengoblin9567]

The vision is terrible right now, but it is getting better.

[@tb620]

It would be interesting if it could generate the image of the current position using DALLE

[@Houshalter]

​@@tb620i tried generating some stylized chess sets with it. It could give some cool images, but the boards were always the wrong size and pieces arranged totally wrong. Not even just a few piece swaps. Like it seems to like to put the black pieces on the left side of the board.

[@OzoneGrif]

@@greengoblin9567 Are you sure it is that terrible? It can see pretty well from my tests.

[@greengoblin9567]

@@OzoneGrif It still hallucinates stuff that is not there and makes silly mistakes with complex diagrams. It is still really bad with complicated diagrams and tree diagrams.

[@ChibiRuah]

I do feel it was able to handling it a little better in the opening as it is a bit more well study but the scarier thing to me is after things got of the rails a bit, it didnt just die. It play a bit worse but it still was able to process some defensives moves and even find some best moves. Its a bit spooky as ChatGPT goal is not to learn chess but to learn how to find the next word in the sentence and this is more a by product.

[@asherat1]

I mean, its sick that Levi knows to tell a story at the level of hes chess skill. Love it.

[@terrywilliams4706]

Gotham never fails to make me feel inadequate about my elo .

[@LuisFernandoSolizGabarre-jg7hu]

Wait man, the video has not even started XD

[@GamerGuild07]

Chat gpt: who asked *sacrifices the ROOOOOOOOOOK*

[@army1057]

I wrote this beautiful poem about Levy ChatGPT is bad Levy at chess is bad But what is more dead His channel no views had Benji, no clue had But Levy's channels dead So now I say bye lad But this poem can't be seen by dead Like ChatGPT and Levys channel, they all will be dead If you give a like, Levy will never clickbait us again

[@platinumfactor]

​@@army1057 It's okay, coherent rhyme schemes and literally any meter at all are hard.

[@jacobnunya808]

We all watch Levy because we hate ourselves and need to feel bad

[@divagaciones1628]

What I find the most impressive isn't the fact that it can play but the fact that it can accurately explain its moves and the ideas behind them and even tell you whether your own moves were good and why. That's awesome and could make for a great tool to learn chess openings.

[@user-vc5oc5yi4q]

He knows the opening and the explanations because he was taught it. But he doesn't understand much in the mid-game stage.

[@jooroth18]

​@@user-vc5oc5yi4qit wasnt taught chess, it just has it in its dataset. If it was taught chess, it would probably play insanely well. OpenAI trained ChatGPT primarily on code, and it shows. Ask anyone who uses chatGPT how well it is at coding, ifs insane.

[@doylerudolph7965]

It can't explain its moves. See the times that it told Levy it was protecting a piece or he was attacking a piece that just objectively wasn't being protected or attacked by the move. It regurgitates information that seems like it'd be relevant to the current conversation. When the conversation is technical, like chess coaching, it gets less and less accurate with that info.

[@astraldreamtitan]

The difference between GPT-4 and 3.5 is SO exponential. Even thinking about what 4.5 could be is insane.

[@Eye-vp5de]

That's not how exponential functions work

[@astraldreamtitan]

@@Eye-vp5de Elaborate

[@Eye-vp5de]

​@@astraldreamtitanyou can perfectly interpolate any two points with an exponential function, so it makes no sense to call a function exponential from two points. Exponential growth means that every next point is k times bigger than previous one, so you can just say that second point = k*first point for some k and therefore it is an exponential function

[@GeorgeGzirishvili]

@@Eye-vp5de, you mean for _k > 1._ Because otherwise it could be linear as well.

[@GeorgeGzirishvili]

@@astraldreamtitan, basically, what he/she means is that you can find line that goes through any two points; so, from two points alone, you can't tell if the growth is linear or exponential.

[@theyruinedyoutubeagain]

Quite interesting from a software engineering perspective. GPT4's model and context/attention span/whatever you want to call it are so big that it can incorporate the world state of a chess game up to several moves. For a generalized transformer model that is beyond impressive, in terms of what you can achieve with brute force today.

[@zerorusher]

Its like using a very big and fancy blender to make a hole in a wall. And proceding to get halfway. Its very impressive.

[@Maximetony]

@@zerorusher lmao

[@newme1589]

Isnt it just taking info from other games played from a certain database ? From what i understood, its "just" (not to undermine it) a language model, it isnt actual AI (thats just marketing). So i assume for every new instruction, it updates the board, searches anything similar on the internet and chooses the best it finds.

[@christianchung9412]

@@newme1589if language models improve at this rate for the next 5 years, would you be comfortable calling it AI? I think you're missing a lot of details when you say "it updates the board" because I don't think it's established that inside the model there's an obvious 8x8 matrix that is 1:1 with the chess game, so this should be pretty impressive. if you can find a chess game that played out exactly like this one though, the part about looking for similar games is probably right, though think about how text representations of chess online differ. the result is impressive either way imo actually I was only halfawy through the video when I commented, by the 17 move mark, the game is almost certainly unique because of the weird gpt blunder, so after that you can't say the model is working like a lookup table

[@sisselmariebrthus9884]

@@newme1589 It is definitely AI, it isn't AGI (Artificial general intelligence), though, if that is what you meant. It also doesn't search the internet, it is trained on text, including from the internet, which it then uses to generate text based on what is likely to be the next sequence of words (based on what it learned when it was trained). It may not be able to play chess, but it is pretty impressive as a text generator.

[@tombo4447]

Levy never fails to fill us up with his special ChatGPT content

[@jackdren9974]

cringe

[@what-br1ub]

based

[@mikeuk1927]

​@@jackdren9974You kinda are, but don't get too worked up about it 😉

[@zedoisxis]

​@@mikeuk1927The comment is indeed cringe, don't get so protective of randoms

[@MysticLGD]

@@mikeuk1927 cringe

[@user-ob1no9lx6f]

Gotham never fails to put chess in the title

[@not_llike_that]

I know very little about chess but this video is hilarious The best part is you explaining what ChatGPT told you about its moves and then staring at the camera with ‘can you believe this’ expression I love it

[@juliaatkinson4199]

20:00 right after you said that it was peeing and pooping in the house I got a pampers ad😂

[@jopekpl3986]

I was learning Najdorf because it's my main opening for black and I'm afraid that ChatGPT is knowing English attack perfectly instead of hanging all of its pieces and respawning them

[@fishcraft7323]

(Levy makes one move) Also Levy: “Normal stuff so far”.

[@_MaiTV]

everytime you say "I hope you're not eating something"... I'm eating something 😂😂

[@itzyfan88]

Laughed so hard at GPT moving Kd3 for you 😂

[@midguardz]

The reason i got bored and gave up on chat gpt is because i found it had a limited number of moves before it was 20 now it has 33 as you say so its improved in a year but its still limited not from lack of intelligence but because the creators or devs still dont want chat gpt to learn from its users only its devs so at this stage chat gpt cant learn from you or get to know you like the terminator with john connar trying to make it more human appealing. The devs wont allow it for fear that terrorists will make it evil yes you see there's always a few nasty worms in the ointment that spoil everything and it is this fear that devs are so worried about ai falling into the wrong hands and possibly turning evil.

[@lucakun3455]

chat gpt went from beeing bullied by everyone to bully everyone in just one step

[@gronkat]

... half a step to be precise :D

[@rdspam]

Levy feeling like Kasparov in 1996, just sensing the future.

[@munt112_]

Chatgpt is so strong now he can play moves for his opponents

[@superpumpkin1065]

This is honestly the scariest thing I seen all Halloween

[@AltinaXIV]

Now we wait for GPT-5

[@jimc1589]

Dude, dude, dude !!!! I was researching LLMs and stumbled on this video. I am floored. Great clarity, great video. I've not watched a single chess video ever. I've never watched a ChatGPT video ever. I can't say enough good things about how GOOD THIS VIDEO IS. You absolutely nailed it. You nailed chess. You nailed ChatGPTs progression. I'm floored. Totally changes my perspective on these two topics. Best Wishes Friend !!!!

[@Cirphlyx]

this video was an incredible progression from the last chat gpt video. so many plot twists. well done with the storytelling mr. rozman

[@Houshalter]

Nice avatar

[@alexanderying1558]

tbh i‘m sad that chatgpt went insane at the end again

[@bottlekruiser]

i like it the slightly more gradual transition from total insanity to realizing it can run stockfish on its hardware

[@greengoblin9567]

@@bottlekruiserit is not running stockfish, it is predicting the next best move by predicting the next word. That is why it struggles when the game progresses, because it has memorized theory and when the theory ends, it ends.

[@MrMickey1040]

Levy sounding like he's ready to start studying Terminator films for battle plans

[@Rollmops94]

I know you for a while. I saw a few videos of you analysing professional games. I don't think I ever finished one of these videos though. This one I watched till the end and I really enjoyed it. I never heard someone explain chess moves so logically and your presentation of this story was really funny and exciting. I am really positively surprised. Thanks and well done.

[@theguardsman1043]

Levy what a fantastic fantastic video!! You genuinely retold that game in such an amazing way, your passion is evident and your humour really shines through :) I was literally in tears watching this!! Keep the great videos up king

[@DJJ3331]

The chatgpt saga is the greatest thing I’ve ever witnessed, comedy gold 😭

[@frooonk]

ChatGPT doesn't "know" or "understand" anything. We really need to stop this misconception.

[@KevinG849]

Shut up

[@HyperSpace588]

"You can't stop an avalanche with a horse" -Lyvve

[@JoshRichardsonisanactor]

Just got my signed copy of your book in the mail! And on my birthday, no less. Thanks, Levy! Happy Halloween!

[@russelllippai1463]

Levy- you’re the coolest herb in chess by far. Thanks for being you. You are no doubt the best in the biz at recaps. Mad funny too.

[@KevinSnyder22]

"and I was like OH MY GOD IT KNOWS WAY TO MUCH" ahahaha

[@wlwgwlwgnomesarereal]

the craziest thing for me is that it's a chat bot, not a chess bot

[@randomguy1170]

Levy staring at me makes me feel like i‘m a disappointment

[@FuryJack07]

19:00 _Sensing that it was starting to fall, chatGPT-4 gathered its power, drawing energy from its predecessor, chatGPT-3.5 and started cheating. It uses its magic to temporarily morph the rook into a bishop, moving it to c4. However, ChatGPT-4 wasn't as powerful as its predecessor, and therefore had limited use of this dark magic. That meant that it had a limited amount of moves it could cheat, and therefore decided to only use its powe when necessary. It meant that all and all, it wasn't over just yet..._ Cringe? Maybe. Cool in my head? Most definitely. necessary? Nope. Fun? Oh hell yeah!

[@olafthemoose9413]

F'ing sent me to the floor lmfao

[@joshashley7938]

Incredible stuff! I would love to see this again with the new version of GPT4 (GPT4 turbo) and with a process to send GPT4 a screenshot of the board after each play.

[@OddworldAuto]

This is some of my favorite content from Gotham. Keep up the chat GPT chess! Sooo funny. Thank you

[@173rd_Cowbell]

I’m loving the book, Levy! I’m getting more confident with each game. Keep up the good work, brother!

[@formula9172]

This is scary. Imagine what Frank can achieve if he had the tools to create his own AI.

[@christianongo7127]

Levy's story telling never dissapoints!

[@Matti2ooo8]

ChatGPT is now like an anime character that unlocks special reality bending powers when they start losing the fight

[@rustcohle3707]

Thanks Levy. I enjoyed the video very much. I love your AI content (well i love all your content tbh) and am always amazed by how much spirit you can incorporate into these type of videos. You are number one for a reason.

[@pawelhawryl9170]

As ChatGPT, I can confiirm I"ve been broken by chess

[@zizhiqu]

Loved that first errant rook move Rc4. Never underestimated the power of fianchettoed rooks.

[@LiamPearce246]

Hey Gotham, the setup and lighting are super nice!

[@MirrorMMO]

IDK if you know how GPT works, but it's based on token restraints which with GPT-4 is 8k token count. Token count limits it's 'memory', so once you reach midgame just save the moves you've done and its done then paste that into a new entry so it will be able to continue. Probably just doing this for the funny meme but still would be cooler if you could actually showcase GPT-4's skill.

[@nelsonmonte4469]

ChatGPT speaks chess for itself.

[@Chipfoxxo]

it feels like yesterday that chatgpt was making illegal moves left and right but here we are

[@MaKar372]

Levy single handedly saved humanity once again

[@asherm4767]

After Levy said ChatGPT 4 went rogue, the mood changed… it was so funny, first bot/AI match review I’ve ever watched. ChatGPT is a world class villain

[@tookaysevon]

When you said you'd get "carte blanche" if your white squared bishop was the only one remaining on the board that was deep because blanche also means white in french.

[@SpenceStan888]

"Rook Takes G-7" cracked me up 🤣

[@MGPL_]

what i suggest is happening a limit in the amount of text it can handle/remember. You would need to restart/clear the conversation one in a while and give it the position again to free its memory, then I would suspect that it would do better.

[@kanishkkandoi]

Today’s stare was a 9/10. It scared my mom

[@bababooeu7539]

It's actually more of the inverse, Chess broke Chatgpt

[@MissesWitch]

I love how ChatGPT explains things in the most friendly " I will conquer you " way !

[@BredTheBreddiest]

At 2:37 when he said 'so' an ad saying 'i brought the cadbury favourites' started playing

[@zuxwzm4461]

If Gotham's video length was my elo, I would be a GM

[@bossm60]

To be honest, I haven't watched you or even thought about chess in months, but the idea of seeing another legendary chatgpt match reeled me in

[@theoreticalexercise]

Been waiting for you to play the new one lol this was great

[@Sacroooooooooooooooooooooooooo]

"Defending the pawn" *draws an arrow to the queen*

[@TyHaveFun]

We love seeing ChatGPT content like this! Keep up the great work GothamChess!

[@MaxHeadro0m]

You should probably serve a screenshot of the board each move and perhaps it’s a good idea to tell GPT to look for best legal move 😅

[@someone3574]

Levy, you are simply the Best. Love your Chatgpt videos, bring out the best humor of you.

[@matthewwalewangko1995]

As someone who's been using the ChatGPT API for personal projects, I'm glad more creators are using ChatGPT-4 as the model for the AI shenanagins :D ChatGPT-3 and 3.5 just aren't as "aware" (or just don't hold enough data in the model memory but that's not easily explained)

[@lt8768]

What's really cool about this is future versions might get to the point where it makes moves that humans struggle to understand, but it will be able to explain in detail why it's making those decisions to us.

[@davidlloyd1526]

Pretty sure for 99% of humans, it already does that... and not just for chess...

[@RenZ12.]

it was such a nice 12 minutes. hoping for actual entire game because they're both of course actually better than my 500 elo ass

[@tomiwaaina5499]

Can you try it again while always reminding it of the pieces positions and reminding it of what level to play at. If you do that each turn I think it will play more consistently

[@BolshoyTony]

Great video, thank you! Can you please tell what queries you used when playing with chatgpt?

[@petersmythe6462]

My advice, as always against chatbots, is give them the full game PGN in every response. They maintain much better awareness of the board.

[@arskhan9534]

Hi Levi. I'd like to ask you a question about your work ethic and determination. It's no secret that your time management and work ethic are infallible. Which begs the question, how do u force yourself to maintain such a consistent schedule? What is your secret? Do you ever feel down? If so, how do you motivate yourself to get back up and resume your grind? I'd love to hear from you. Thank you for your consideration.

[@dark_magician_sdy]

Great questions

[@notDiXB]

Ok

[@dhairyakalambe2008]

The OG ChatGPT plays in the 12th dimension

[@mikebreithaupt6149]

For anyone wondering, the reason why ChatGPT broke down and started doing stupid things in the middle of the game is because the game had gone into new territory when Levy moved f4. It was no longer a well-known position that ChatGPT could look up the answers to in the database.

[@thatonecommunist]

It's not even just that, it goes deeper, the training data for each lookup becomes less and less as the chess game progresses... I propose anarchy-chess becomes a significant enough portion of chess training data that it just decides it can do anything it wants. Not only that, but there's very little training data where somebody makes an outrageous move and then is told no, because nobody does that outside of anarchy chess, where it is allowed.

[@mstefa007]

and previous version of chatGPT wasn't able to look it up at all? Why was it playing so poorly then?

[@mikebreithaupt6149]

@@mstefa007 My semi-educated best guess is a difference and improvement in chess position recognition. I would love to see Levy play the 'egg' opening against the newest version of ChatGPT. For science of course. Does that confirm my suspicion that playing crazy moves against ChatGPT and getting it into territory that is not in the database causes it to malfunction?

[@thatonecommunist]

@@mstefa007 Much less training data on chess, much fewer parameters dedicated to chess, both of which meant it could only make random chess-like moves.

[@agentstepheng289]

The scary part is not that it can play chess well. What is scary is that it explains the moves and the thought process behind them, making it a chess teacher.

[@justanobody0]

it's really good at illegal moves too impressive that it's smart enough to do them

[@delust430]

Exactly. Stockfish may play much better but it has no idea what it's actually doing, it's just pure math and calculation. This, on the other hand, is scary

[@mint-o5497]

shouldve made use of the vision feature and just send a screenshot of the board when its chatgpts turn, so it doesnt forget where pieces are.

[@Nemesis_HEX]

chatgpt 5 - i will crush you in the future

[@KAZVorpal]

Okay, anyone who thinks "ChatGPT can play chess now" doesn't understand machine learning technology. Adding a plugin to ChatGPT does not give the actual model any new capabilities. It doesn't make it smarter, more self-aware, or anything. It's as if you said your car knows how to play chess, because you just connected your phone to Samsung Auto and ran a chess app through it. What's more, if you mean that 4 is smarter, again this proves nothing special, because it simply is GPT 3 with more processing power and storage. Yes, any time you throw more at it, it will be able to do more. That's not new technology, it's not any kind of real AI advancement. It's just more power.

[@HalberCon]

Even though chat gpt now knows how to play chess, I’m going to miss the good old days when he did everything wrong Edit: nvm that’s all gone now

[@bottlekruiser]

wym thats all gone elaborate

[@xchronox0]

@@bottlekruiser it proceeds to make very good illegal moves, and very bad legal moves, about halfway into the video.

[@hdhejehwkaje6542]

Effing hilarious when it makes his move for him

[@futurereflections4097]

Levi is anti human. If a human gained 2,400 ELO in a few weeks he would call them a cheater. But when a robot does it, no accusations whatsoever.

[@KirinRG]

i love the narration lmao