It feels like they're really focusing on overstating how confusing and weird it is that an LLM can write code but not play games very well, rather than just explaining it.
Code is text. LLMs are text input/output machines.
Game input/output is not at all text.
LLMs can certainly reason about games with a simple/explicit enough domain (try a risk tournament where models can talk to each other between turns!)
I actually really miss all the research being done on having (reinforcement learning) AIs beat Atari games and the like. Or the one that stopped at a TV playing random images instead of continuing through the level. Has there been any progress in that field? It seems like LLMs came around and all the projects stopped completely.
I wonder if you paired a few different types of AI together, an LLM agent might be good at strategizing -. E.g. building a strategy on how to handle a scenario. But, it would need to know the entire game manual basically. Then it would pass the stratrgy to a better AI in some way. But it might not be needed if the better gaming AI can just do that part too already.
I know someone who tried the "aibot plays pokemon" thing...
From what I saw, even if you frame advance every single frame, they still don't seem to grasp the concept of "I need to hold down this button for a few frames until x happens"...
There's no concept of time, just a never ending state machine thats constantly changing state.
It matters a lot because it's a real solution for external bots that plays more "fairly" especially in older games. It also allows to test games autonomously, which is huge if we are talking about automated programming.
Imagine if you can bring those AI players to CS 1.6.
LLMs are the wrong tool for video games. There have been plenty of successful non-LLM AIs that have been trained with reinforcement learning to play games.
If you want to implement actual bots inside the game, then you want to use explicit logic instead of inferred logic. It's much more efficient and easier to debug.
> This brings us to what seems like a contradiction. LLMs are bad at playing games. Yet at the same time, they’re improving rapidly at coding, a skill set that can be used to create a game. How do these facts fit together?
> Togelius: It’s super weird.
...No, it's really not.
They're language models. Code is a language. "Playing a game well" is not. One can, hypothetically, encode game inputs in such a way that it seems kinda-sorta like a language, but it has none of the same kinds of structures that languages—both human and programming—do.
The only way one can think this is strange is if one thinks of LLMs' ability to code rudimentary games as being due to a deeper understanding of games, rather than due to game code being well-represented in their training data.
Yet LLMs can play chess and have a "mental" representation of the chessboard.
If LLMs get better but do not progress at playing games when not specifically trained on it it seems to point to a generalisation failure, a limitation that would prevent LLMs to ever achieve AGI, I do not know if that is weird but it seems that for now nobody really knows if they can achieve AGI or not. Perhaps some emergent behavior will arise after more scaling.
To me it's only totally unsurprising if you are 100% certain that LLMs will never reach AGI (like LeCun thinks for example).
Yea it’s wild watching so many smart people convince themselves that LLMs are general purpose AIs. Don’t get me wrong they are incredibly powerful tools. However being surprised that text models cannot play video games particularly well is like being surprised weather models cannot.
There was good progress in training neural networks to play video games.
Unfortunately it doesn't seem to fit in some people's context because it was a few years ago.
Kind reminder: there is "AI" beyond LLMs.
OpenAI's Dota 2 adventures were super hype back in the days.
Because they’re large language models. Language doesn’t map onto gameplay.
Choose another “AI” technology and give it about go.
It feels like they're really focusing on overstating how confusing and weird it is that an LLM can write code but not play games very well, rather than just explaining it.
Code is text. LLMs are text input/output machines.
Game input/output is not at all text.
LLMs can certainly reason about games with a simple/explicit enough domain (try a risk tournament where models can talk to each other between turns!)
I actually really miss all the research being done on having (reinforcement learning) AIs beat Atari games and the like. Or the one that stopped at a TV playing random images instead of continuing through the level. Has there been any progress in that field? It seems like LLMs came around and all the projects stopped completely.
Why is a language model bad at video games? I think the answer is stated in the question itself.
I wonder if you paired a few different types of AI together, an LLM agent might be good at strategizing -. E.g. building a strategy on how to handle a scenario. But, it would need to know the entire game manual basically. Then it would pass the stratrgy to a better AI in some way. But it might not be needed if the better gaming AI can just do that part too already.
I admit I know nothing about this though.
GOAP is a better tool.
I wonder if they would be good at text-based games.
Its almost like the Large Language Model has trouble with things that arent Language, such as realtime controller input and video output from a game
I know someone who tried the "aibot plays pokemon" thing...
From what I saw, even if you frame advance every single frame, they still don't seem to grasp the concept of "I need to hold down this button for a few frames until x happens"...
There's no concept of time, just a never ending state machine thats constantly changing state.
Video games are made to entertain humans, so does it really matter whether LLMs are good at playing them?
It matters a lot because it's a real solution for external bots that plays more "fairly" especially in older games. It also allows to test games autonomously, which is huge if we are talking about automated programming.
Imagine if you can bring those AI players to CS 1.6.
LLMs are the wrong tool for video games. There have been plenty of successful non-LLM AIs that have been trained with reinforcement learning to play games.
If you want to implement actual bots inside the game, then you want to use explicit logic instead of inferred logic. It's much more efficient and easier to debug.
If you want to create Bots for an existing game, which doesn't have its own pre-programmed bots, then you should look at other types of AI. See https://www.geeksforgeeks.org/deep-learning/reinforcement-le...
The headshot/spin bots didn't need ai, all they had to do was ask the server where you were standing, and teleported to your location.
> This brings us to what seems like a contradiction. LLMs are bad at playing games. Yet at the same time, they’re improving rapidly at coding, a skill set that can be used to create a game. How do these facts fit together?
> Togelius: It’s super weird.
...No, it's really not.
They're language models. Code is a language. "Playing a game well" is not. One can, hypothetically, encode game inputs in such a way that it seems kinda-sorta like a language, but it has none of the same kinds of structures that languages—both human and programming—do.
The only way one can think this is strange is if one thinks of LLMs' ability to code rudimentary games as being due to a deeper understanding of games, rather than due to game code being well-represented in their training data.
Yet LLMs can play chess and have a "mental" representation of the chessboard.
If LLMs get better but do not progress at playing games when not specifically trained on it it seems to point to a generalisation failure, a limitation that would prevent LLMs to ever achieve AGI, I do not know if that is weird but it seems that for now nobody really knows if they can achieve AGI or not. Perhaps some emergent behavior will arise after more scaling.
To me it's only totally unsurprising if you are 100% certain that LLMs will never reach AGI (like LeCun thinks for example).
[delayed]
Yea it’s wild watching so many smart people convince themselves that LLMs are general purpose AIs. Don’t get me wrong they are incredibly powerful tools. However being surprised that text models cannot play video games particularly well is like being surprised weather models cannot.
cough JEPA cough