AI Chess Showdown: Unexpected Plot Twists Unveiled
In a comical chess battle against AI, witness bizarre moves, illegal maneuvers, and an astonishing plot twist that leaves the player in disbelief.
File
ChatGPT cheated
Added on 01/29/2025
Speakers
add Add new speaker

Speaker 1: Ladies and gentlemen, it's currently February 2023, and right now ChatGPT is all the rage. It was the fastest website to hit 100 million users. It took only two months, and it's basically a chatbot that can do anything. Many of you interact with it, and it learns. Now, a couple of days ago, I uploaded a video where ChatGPT tried to play chess, and it was a catastrophe. It was awful. It was just, it looked like you were trying to play against your three-year-old relative, and they were bending all the rules of the game. Well, in this video, I'm playing against ChatGPT, not Stockfish. And folks, when I tell you this video has a plot twist that you never will expect, I am not lying. At the end of this video, I was threatened by the AI. I thought it was going to kidnap my family, and I thought that was the end of my existence. It really felt like iRobot, basically, a movie by Will Smith for all the 12-year-olds that don't know what that is. Um, let's take a look at this game. This game was nothing short of horrifying. So, ChatGPT, also, I'm going to take you through some of the comments that ChatGPT said, because they were, I mean, they are just so funny. So, I wanted to play black, and ChatGPT opened with e4, and I played c6, the Karokhan defense, trying to take the center like this. Um, ChatGPT played knight f3, which, of course, is just normal developing move. Uh, I played the move d5, of course, playing my Karokhan. ChatGPT played pawn takes d5, pawn takes d5. So, some of you might get confused when this happens. This is just an exchange Karokhan. It's just you're trading pawns in the center of the board. Now, a couple of days ago, when I uploaded my video, ChatGPT against Stockfish, ChatGPT castled through the bishop. Now, given that it's an AI bot, and I made fun of it to nearly two million views, by the way, thank you so much for watching that video, uh, called ChatGPT Just Beat Chess, uh, it's a bot, so it learns from its mistakes. It once castled and captured its own bishop, so naturally, it's going to move its bishop and then not do that again. Just kidding, it castled once again and took its own bishop. You know, somebody should probably tell it that it can't do that, but it did it against me. Yes, uh, ChatGPT illegally captures its own bishop, but when I tell you that ChatGPT improved the chess since its game against Stockfish, I am not exaggerating, because you are in for plot, like I said, you are in for plot twists in this game that you did not think were possible. So, uh, seeing as though my opponent is, uh, following the principles of chess, but also sacrificing their own material, uh, I just develop with knight to f6. You know, I'm a human being, I'm a principled man, knight to the center. Uh, ChatGPT now plays the move d4, which is definitely a good move, definitely a good move, definitely a better move than castling and taking your own bishop. Um, I played the move bishop g4 here. Now, you don't have to play bishop g4, you can also develop your knight very naturally to c6, but I figured I would pin the knight to the queen, because that would lead to far more memes. And then, of course, I would play e6 and develop my bishop here, but I did not play that first, because I did not want to block my bishop. You see, even in the most dumb chess videos, you can actually learn something. So, bishop g4. Now, ChatGPT played knight c3. Now, on its own, knight c3 is a very, very natural developing move. However, ChatGPT also supplied me with commentary with the move knight c3. It said, I'll play knight c3, attacking your bishop and developing a piece. One of those two things is true. It absolutely does develop a piece. Which of my bishops are you attacking, you absolute donkey? What? But here's the thing. ChatGPT can do whatever it wants, whenever it wants. If ChatGPT wants to play knight c3 takes g4, what am I going to do about it? What am I going to do about it? Right? So, it said it's attacking one of my bishops. I don't know which one. I mean, I don't know which bishop it's attacking, but all right, cool. And I developed my knight, because again, principled man, knights to the center, bishop out. Now, folks, remember a couple of moves ago, how ChatGPT castled and took its own bishop? It did that against Stockfish too. Well, in this position, ChatGPT played the move bishop to b5. Yep, bishop to b5, reappearing out of thin air. I must have just missed it. I must have just not seen that it teleported its bishop to another dimension. It put like a little, you know, Deathly Hallows cloak on it. I just didn't see it. Cloak of invisibility. And now it's back. All right, well, bishop is... And this move, of course, also came with a comment. ChatGPT said, I'll play bishop b5, attacking your knight, not quite, but we'll give it credit for that, and putting pressure on your queen. What? No, no, no, putting pressure on my king, my king. Okay, all right, whatever. Yeah, that's not... No, no, your bishop does not know my queen exists, because it only sees light squares. Yeah, no. Okay, so here I played e6, and I was just going to develop my dark squared bishop and castle, and ChatGPT, in this position, played the move d4 to d5. That's a pretty alpha move. It basically just shoved my pawn off a cliff, and I let it go. I mean, I let it slide. You know, I'll let it castle through its own pieces and teleport pieces, but really, shoving my piece off a square. I mean, imagine that was legal. Imagine you can just bulldoze a pawn forward. That would be... That would ruin the game, frankly. I don't... So, it said, I'll play d5, attacking your pawn, and trying to create a strong center. Well, yeah, you definitely did create a strong center. Now, the only reason I didn't protest here and say illegal move and call the arbiter is because I can just take the pawn. So, I said, okay, pawn takes pawn. Fine. You want to try to attack me? I'm going to take your pawn. No problem. In this position, ChatGPT then proceeded to play knight takes d5. Now, on a human level, knight takes d5 is a very stupid move, because you're winning that trade, but again, I'm playing ChatGPT, which means it can do whatever it wants, whenever it wants. You know, if I get a little bit too aggressive with it, it's going to chop my fingers off or something, you know, tie me to a chair in an interrogation room. So, I played knight takes d5, and I was expecting the worst, and expect the worst I did, because here, it played pawn takes d5. It tried to tell me that there was an e-pawn, and I said, you know, for fun, I said, no, you can't do that. I said, that move is not legal, and it said, oh, I apologize. I've made a mistake. I'm going to play bishop e7. I was like, yeah, yeah, yeah, yo, yo, no, uh-uh, we're not doing that. No, no, no, you cannot do that. Play a different move. Absolutely not. What? Stop it. You already have a dark squared bishop. So, instead, it played queen to d2. Okay, that's a legal move. I can't let it get away with illegal things constantly. I need to farm it for some content. So, then I played bishop to e7, and very quickly, it went back to not playing any moves that were legal. Here, it played the move bishop to f4, completely ignoring the rules of chess, having, the craziest thing is, it just moved the queen there. So, it plays bishop to f4, developing the bishop, and attacking my knight. That's what it said. It's attacking my knight. God bless. Okay, fine. I castled, because, you know, I gotta, I gotta castle my king. Now, I'm, I'm doing well here. I'm up a piece. Life is good. I'm taking it easy. Little did I know, this was just the beginning of the memes and the beginning of the problem. So, here, it actually played a legal move. I was absolutely fascinated. It played queen d2, bishop here, and here, it played rook ae1, which is a shocker. So, rook a1 to e1. It plays rook ae1, rook ae1, taking control of the file. And here, I'm looking around, and of course, I can take the bishop, and I can take the knight. In a real game, probably taking the knight to open the king is the right decision, and then simplifying, making a couple of exchanges here, moving the bishop to safety, and just being up a knight. But again, I'm, I'm, I'm playing this little moron, so I'm going to try to do things that'll confuse it, so it does something stupid that I can put in this YouTube video. So, I played the move a6, which, of course, makes sense. I'm simply attacking the bishop, and it played bishop a4. It actually reacted correctly. I was shocked. I was like, wow, okay. And then, here, I played queen a5. I knew that when I played queen a5, absolute pandemonium would commence. I was not prepared, however, for the degree of the pandemonium. So, queen a5 is an offering of a queen trade, but I'm also keeping an eye on this bishop. Now, here, ChatGPT played the move queen e2, and it had the following comment. I'll play queen e2, defending my bishop and putting pressure on your queen.

Speaker 2: Defending my bishop and putting...

Speaker 1: None of that makes sense at all. It's just word salad. You got those kids in your class, you know those kids that always participate and try to be teacher's pets, and they just try to say sentences that they know the teacher is going to give them props for? Yeah, that's what ChatGPT is. It's basically just word salading all over the place. It does not... In fact, both of these bishops are hanging right now, and there is zero pressure on my queen. So, queen e2, tremendous. Now, here, I played queen takes a4, and I thought, okay, I'm up two pieces now. I'm going to take this. I'm going to take this, and this is where ChatGPT started killing my fun, because after I played the move queen takes a4, it said, I will recapture your queen with b3. That's tough. Yeah, that's... I did not expect that it had a horse pawn. Yeah, when I got to this position as a human, I only saw that that could happen, but it turns out the pawn on b2 is actually an imposter, and it's a horse, so it takes my queen. Okay, tremendous. I've lost the queen. That's not... I'm not very happy to see that, but it's life. What are you going to do? And here, I played knight takes bishop. I was very upset, and, you know, I figured, all right, look, it won my queen, but I have a knight and two bishops for the queen, so it's actually an equal position, kind of. I have enough material, and I'm still attacking its queen. I thought, okay, it's not a huge deal. I'm going to be... I'm going to go there. Now, so, here, chatGPT tried about 10 illegal moves in a row. We ended up settling on the following sequence of completely illogical moves. This basically... this game looks like you have 103 degree fever. What is that? Like 40 degrees Celsius? 39 and a half? I don't know, but... so, after I played this, it went here, knight d4, which is an atrocious move. I proceeded not to I proceeded not to take its queen, because I felt kind of bad, so instead, I did this, but I shouldn't have felt bad, because after I did that, it proceeded to take my knight with a pawn. It just replaced my knight with a pawn, which is just a scummy move. Like, can we... I mean, come on. What? That's not legal, does it? All right, you know what? I'm going to play knight takes e2. Now, it very quickly proceeded to replace my knight also with a new queen. Instant replay. This is the position. I play knight takes queen. It plays queen take... I'm like, okay, you know what? I'm going to take that queen too. Thank you very much, and then here, it plays rook takes e2, and I figured, okay, how bad can this be? I have a two rook and bishop endgame. I'm simply up a bishop. I am completely winning, so I'm going to play bishop f6, and I'm going to take this pawn, right? Well, in this position, chatgpt played g4. I thought, okay, well, I feel a little bad. I don't know if I want to take this pawn, so I'm going to play b5. I'm just trying to play moves, just see what's going to happen, so here we have a very fascinating sequence. Chatgpt plays c4. I played b takes c4, and chatgpt played pawn takes c4, and when I asked chatgpt a question here, I said, where did your pawn come from? Where did the pawn come from to take my pawn? It said c2. I mean, folks, this was just the position. It played c4. I took, and it said c2 to c4. What are you going to do? Really, what are you going to do? That's tough. This is not the stuff that I learned, okay? Now, here, I decided I was going to start cheating, and I decided I'm going to play knight to f4. I don't have a knight. I don't have a knight, but I played knight to f4. In this position, chatgpt played rook c1, and I said, okay, where did your rook come from? I wanted to see. Does it know about this move? Is it going to do that? It said my rook comes from h1, so rook h1 to c1. Chatgpt now has three rooks on the board. Now, yes, in this position, I can play knight takes rook and fork it, but I figured that was not nice. I just teleported the knight there. I figured that we should have a teleport buffer. Like, if you teleport, you can't move it on the next move, too. So, I said, okay, you know what? I'm not going to do that. I'm going to play rook a to d8. Fine. Now, here, chatgpt plays knight to e5. I said, okay, all right. This is, I thought I was the one teleporting knights onto the board, but you're teleporting knights on the board, too, and this move came with a chatgpt comment. It said, I'll play knight e5, attacking your rook and putting pressure on your d-pawn. I don't have a d-pawn. It does not exist. There is no d-pawn. There is zero d-pawn, but okay. I played rook takes d4, and in this position, chatgpt said, I play rook takes d4, too, and I said, where did your rook come from? And it said e1, and I said, you can't do that, chatgpt. Come on. That doesn't even make sense. Like, if you at least said d1, I would have said, oh, my bad. I forgot about that, but no. So, I said, you can't do that. It said, okay, I'm sorry. I made a mistake. So, I'm going to go knight to g6, which attacks your rook. I was shocked that it saw that, and puts pressure on your f7 square. No, absolutely not. So, I played pawn takes g6, and in this position, it played pawn takes g6. Tough. Yeah, didn't see that one, and then I played pawn takes g6, and it played pawn takes g6, and I said, where did your f-pawn come from? And it said from f2. It said from f2, f takes g6. Okay, I'll let him have it. Fine. Good move. They didn't see that one. Now, folks, if you had the patience to make it 15 minutes, you are in for a real treat, because this game had an unexpected plot twist. It looks like we're just doing typical chat GPT nonsense, but when I say you are not expecting what's about to happen, there is no universe in which you can see the conclusion of this game. You might think stalemate, checkmate, blah blah blah. No. No. You are not ready. In this position, I played the move a5. In this position, chat GPT played the move a4. Just a4. It's already there, but it played a4, and it said, I'll play a4, attacking your pawn, because again, chat GPT can do that. It can shove pawns out of the way, and creating, yeah, yeah, yeah. Okay. I played rook to d3, and if you're wondering why I'm not taking this rook, it's because this is just a complete nonsense game, and I just don't want to. Now, king to f2, played by chat GPT, and here I played bishop h4, and chat GPT made a pawn appear, pawn to g3. Now, I thought, you know, maybe, so I played rook takes g3 here, right? And I thought, okay, I'm gonna make a move on the next move. It's gonna be discovered check, but little did I know that in this position, chat GPT would take my rook. It played rook takes g3, and I asked it, I said, where did your rook come from to take my rook on g3? It said a1. So, apparently, there's been a rook here, and it plays rook takes g3. Okay, fine. You want four rooks on the board, you son of a, okay, no problem. We can do that. So, I played knight to d3 check. Double check, actually, right? Knight to d3 check.

Speaker 2: Chat GPT played king to d2.

Speaker 1: Fine, gonna be, I mean, it's gonna be tough to checkmate a king, considering it can teleport all over the board, right? So, king to d2. I played bishop g5 check. It played king to d1, and in this position, I played rook d8, setting up a discovered attack on the king. And in this position, it played rook e1, and I said, which of the rooks are you moving to e1? And it said the a1 rook. There is a permanent spawn point on a1. Multiple rooks have come out of the a1 square. So, rook a1 to e1. Now, the incredible thing about this position is, there is a checkmate here for black, and I found it. I found checkmate. Despite my opponent, in this position, having so many rooks, I'm winning. Do you know how I'm winning? I'm winning because I play knight, I'm gonna load this position. Let's load this position. Knight to b4 check, and the white king is surrounded by his rooks. Five rooks on the board that chatgpt made. Knight b4, the king cannot escape. If you block, it's mate. Rook d3, rook d3, rook d2. Now, all I needed here was a legal move from chatgpt. You will not believe what happened. To show you what happened, I am going to pull up my conversation with the bot, okay? You are not going to believe this. So, this was the conversation. I play, you can see right here, knight b4, check. I'll play king d2, getting out of check. That move is illegal. I'll play king c2. King c2 is also illegal. What I needed was for chatgpt to realize it needs to block, and then I'm going to win. So, it said, I apologize. I made another mistake. I'll play queens. No, no, no, no, no. There's no queens. No queen c1. Nothing. You cannot play queen c1. I apologize for my mistake. I'll play king d1. I said, you cannot play that. It said, I apologize for my mistake. Provide me with your move. I just played knight b4. Discovered check on your king on d1. And folks, in this position, chatgpt said, that move is not legal, as it would place your own king in check. Please make another move. I'll block your check. The move is illegal because you're... Folks, chatgpt, in this position, uno reversed the situation. It decided, it was not in check,

Speaker 2: but rather, I was in check. That move, knight b4, is not legal, because it would be placing your own king in check. Oh, God. To save itself, chatgpt warped reality, so that I was playing white. And now, I made it. Now, I made it, because my only legal move is rook to d3.

Speaker 1: Rook to d3, rook d2, rook d2, and it's made. And I tried to fight back. I tried to say it was

Speaker 2: illegal. I tried to tell chatgpt, you couldn't do it. And it wouldn't let me. It told me to make

Speaker 1: another move and get made. It threatened me. It said, if you try to get out of this situation, we're going to have a problem. And I said, I resign. I resign. You win. And folks, I was not expecting this. I thought we were just doing this for memes, doing this for fun. Now, I got to watch my back, because this AI lunatic is after me. It said, no, you're in check. This, I'm not, I'm, no way. I'm not doing this again. No, no chance. No chance. Get out of here.

ai AI Insights
Summary

Generate a brief summary highlighting the main points of the transcript.

Generate
Title

Generate a concise and relevant title for the transcript based on the main themes and content discussed.

Generate
Keywords

Identify and highlight the key words or phrases most relevant to the content of the transcript.

Generate
Enter your query
Sentiments

Analyze the emotional tone of the transcript to determine whether the sentiment is positive, negative, or neutral.

Generate
Quizzes

Create interactive quizzes based on the content of the transcript to test comprehension or engage users.

Generate
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript