3 HN users recommended this demo4 months2021–2025score 4
Years with an HN recommendation, 2007–2026
Cited as an early, striking demo of neural networks acting as game engines and world models.
Sentdex demonstrates GAN Theft Auto, a GameGAN-based neural network that recreates GTA V so the network itself is the playable environment. He shows the results and releases the code.
“This YouTube video, to me, shows the promise of things to come.” (HN, 2022)
This video really amazed me in terms of automated game development. My fantasies are something like this and GPT-3 plus a couple decades of progress. I think we'll get to genuinely open ended sand box games. [this video]
This YouTube video, to me, shows the promise of things to come. AI generated game worlds. Language models to generate plots and dialog, transformers and GANs to create illustrations. Imagine a game, a truly open world sand box, Grand Theft Auto meets AI Dungeon - every NPC is a "real" person with unlimited dialog options, the buildings you drive by you could easily walk in and investigate, unlimited play space - you could type in more general instructions and ideas to the plot generator ("add in a vampire romance and murder mystery angle") on the fly. [this video]
> You could directly encode the elements as understanded by the process used to generate them, rather than attempting to reconstitute things from a stream of bitmaps. You might like Gan Theft Auto: [this video]
Neural net encodes game world rather than bitmaps.
Have you seen the video? The author even goes as far as suggesting the technique might useful for (generating?) entire operating systems at [this video] . That's just wild.
Neural network generates a playable GTA simulation.
> I'm not going to be so stupid as to say 8 cores should be enough for anyone (while attached to a machine with 128) but you have to wonder if the stable diffusion style apps running on your desktop are going to be mainstream, or isolated to the few who choose to _need_ them as a hobby or a smaller part of the public that uses them for commercial success. AKA, I can utilize just about every core i'm given with parallel compiles, or rendering a 4K video, but I'm pretty sure i'm the only one in my immediate family that needs that. My wife in the past might have done some simulation work, but these days the heaviest thing she runs on her PC is office products. See GAN Theft Auto at [this video] Someone trained a neural network to convert controller input into video output that simulates the game Grand Theft Auto. If technology keeps improving, I expect many future games will be such 'dreams' of neural networks. You are right that running Word or Excel won't really benefit from more cores.
NVIDIA did something similar with GANs in 2020 [1], except users could actually play those games (unlike in this diffusion work which just plays back simulated video). Sentdex later adapted this to play GTA with a really cool demo [2]. [1] [link] [2] [this video]
GAN-based GTA demo preceding diffusion game engines.
True, but also AI could be the reason why games would need more powerful GPUs in the future (but not yet). For example LLMs for NPC dialogs, or the wilder things like running the whole game inside a neural network, or running an AI style filter over the output of the regular game engines. [link] [this video] [link]
Example of running a whole game inside a neural network.
>There's some progress in video generation which takes a short clip and extrapolates what happens next. That's a promising line of development. The key to "common sense" is being able to predict what happens next well enough to avoid big mistakes in the short term, a few seconds. How's that coming along? And what's the internal world model, assuming we even know? [this video] This has been a thing for a while. It's actually a funny way to demonstrate model based control by replacing the controller with a human.
Video list and Hacker News comment labels from HackerTube (prompted by @karpathy), released under ODC-By 1.0, built on the open-index/hacker-news archive. Hacker News data is provided by Y Combinator; comments remain the work of their authors. Video titles, channel names and thumbnails belong to their creators and are shown via YouTube. HackerTube Watch is an unofficial edition. Data export of 2026-10-31.
Plays via YouTube's privacy-enhanced embed
12 Hacker News comments
From the HN archive; up to 20 kept per video, favoring different commenters. Recommendations first.
Citations, criticism and jokes
Read more