In case it's not obvious, this isn't a model, this is python scripts that call out to models (code not available).
It's mostly "did you know you can attach an LLM to a PCG system?", but there is one idea here you don't see much of: an image model performs the composition (which image models are really good at), and then you extract the objects into 3d via things like SAM3D before placing them in the world, which is pretty interesting.
The rest is standard stuff you'll find in your favorite PCG system/game engine. Still, think most people don't realize how good LLMs are at 3D these days, especially Fable 5 (which this project predates).
That’s exactly how I’ve used tools like Tripo and Meshy.
Use any image generation tool (or any image) and have a multimodal LLM like Gemini extract individual parts of the image “isolate with a transparent background” - then you can use those images for image-to-3d in those tools.
Don’t forget to use low or “smart” poly features otherwise you get too many vertices to the point you can’t performantly raycast etc.
Side tangent: Smart Poly is often a paid feature, but you used to be able to steal it for free. Because in order to preview it, you download the GLB in the Network tab.
Some 3D AI providers still leak like that, but Meshy came up with their own binary format that splits the stream up into separate files that are assembled with a response it only gets if you have Premium.
Thought it was cool technology - and wonder if browser games in general are doing stuff like that to protect 3D assets from being easily ripped from the Network tab.
> Don’t forget to use low or “smart” poly features otherwise you get too many vertices to the point you can’t performantly raycast etc.
I've found it's actually best to do maximal quality generation and decimate/meshopt to desired budget.
If you're raycasting against your raw art you're doing it wrong anyway, none of the big players do this precisely for the reason you mentioned. And beyond physics meshes, if you're doing anything where performance is a problem you probably want multiple LODs so you have to do this work anyway.
These days you can just ask your agent to build you a LOD pipeline and forget about it rather than sacrificing the art; Claude knows Blender.
This kind of content generation can be quite cool to enable new game ideas for indie developers that were only possible with AAA before. But it also makes it hard to gauge the amount of human work that went into the game. When you see an elegantly designed building, environment or easter egg in the game, was that really made by a human (or even prompted by a human), or did the AI generate it autonomously as part of a much larger generation step?
In both the Autumn and Winter examples in the hero images, the algo seemed to have placed buildings on the water in the foreground
In the summer example in the hero images, the building placement + small pockets of water on the left looks odd and low attention to detail. A similar poor quality result as if an uncaring human used a scatter brush...
Curious if the examples are cherry-picked and by how much, or if this is one-shoted
I am sure they are cherry picked. No one would spend time trying to publish results like this without picking the best results they possibly could to showcase.
It looks impressive. However open worlds are at their best with hand placed details and environmental storytelling. Look at Skyrim/Cyberpunk vs. Starfield where most of the world is proc gen’d.
I get that you run this, and then edit it. But the generated villages just aren’t interesting in my opinion. It’s probably great for tencent’s market where you’re mass producing gacha style games, but I don’t think worlds made in this fashion will scratch the open world itch like the best open world games out there.
I'd actually be curious to read an award winning ai book, all the ai books/interactive story generators I've tried have been... Ok at best... Do you know what novels have been written by ai and award winning?
Japanese isekai might not be your cup of tea, but one example is [1] which was way back in Jan, and models have improved a lot since then.
My sneaking suspicion is that there are a lot more cases than get reported, because authors are disincentivized to disclose AI usage, and readers by and large feel icky about reading AI work for reasons unrelated to quality.
I can totally see a weird future where a lot of what people like is AI generated, and everyone lies to each other about it. So it's hard to do research on this phenomenon.
Interesting prompt to world pipeline. I wonder if there can be more style applied than just the "MMO cartoony" style? Like medieval Witcher or realistic Cyberpunk buildings?
I'm sure the answer is yes, even now there are pretty good model generators out there that can do realistic models. Experimental things like this just tend to go with cartoony styles since mistakes and lower quality (read: faster generation) can be chalked up to "lol", whereas realistic styles tend not to be forgiving with art mismatches and shading mistakes.
In modern polished games, the art team spends a lot of time tweaking things to make sure it looks great and coherent. Most people are not willing to have their agent run for a month to do this work. I don't know if it's been tried yet but I would be really interested to see the result.
Yeah any game ready asset has to be low poly. Even if it has 4k textures the polygon (vertex) count must be low to do anything meaningful (raycast, load many models).
Usually a combination of distance and frustum culling (breaking the model down into several that only load when you need to see it) with 4k or 8k textures can make a low poly game ready asset look very high quality
In case it's not obvious, this isn't a model, this is python scripts that call out to models (code not available).
It's mostly "did you know you can attach an LLM to a PCG system?", but there is one idea here you don't see much of: an image model performs the composition (which image models are really good at), and then you extract the objects into 3d via things like SAM3D before placing them in the world, which is pretty interesting.
The rest is standard stuff you'll find in your favorite PCG system/game engine. Still, think most people don't realize how good LLMs are at 3D these days, especially Fable 5 (which this project predates).
That’s exactly how I’ve used tools like Tripo and Meshy.
Use any image generation tool (or any image) and have a multimodal LLM like Gemini extract individual parts of the image “isolate with a transparent background” - then you can use those images for image-to-3d in those tools.
Don’t forget to use low or “smart” poly features otherwise you get too many vertices to the point you can’t performantly raycast etc.
But yeah - it’s there.
Side tangent: Smart Poly is often a paid feature, but you used to be able to steal it for free. Because in order to preview it, you download the GLB in the Network tab.
Some 3D AI providers still leak like that, but Meshy came up with their own binary format that splits the stream up into separate files that are assembled with a response it only gets if you have Premium.
Thought it was cool technology - and wonder if browser games in general are doing stuff like that to protect 3D assets from being easily ripped from the Network tab.
> Don’t forget to use low or “smart” poly features otherwise you get too many vertices to the point you can’t performantly raycast etc.
I've found it's actually best to do maximal quality generation and decimate/meshopt to desired budget.
If you're raycasting against your raw art you're doing it wrong anyway, none of the big players do this precisely for the reason you mentioned. And beyond physics meshes, if you're doing anything where performance is a problem you probably want multiple LODs so you have to do this work anyway.
These days you can just ask your agent to build you a LOD pipeline and forget about it rather than sacrificing the art; Claude knows Blender.
This kind of content generation can be quite cool to enable new game ideas for indie developers that were only possible with AAA before. But it also makes it hard to gauge the amount of human work that went into the game. When you see an elegantly designed building, environment or easter egg in the game, was that really made by a human (or even prompted by a human), or did the AI generate it autonomously as part of a much larger generation step?
> When you see an elegantly designed building ...was that really made by a human or did the AI generate it as part of a much larger generation step?
Does it matter? Why?
In both the Autumn and Winter examples in the hero images, the algo seemed to have placed buildings on the water in the foreground
In the summer example in the hero images, the building placement + small pockets of water on the left looks odd and low attention to detail. A similar poor quality result as if an uncaring human used a scatter brush...
Curious if the examples are cherry-picked and by how much, or if this is one-shoted
In this competitive world, you need to integrate the cold plunge deeply with your lifestyle to stay ahead.
I am sure they are cherry picked. No one would spend time trying to publish results like this without picking the best results they possibly could to showcase.
It looks impressive. However open worlds are at their best with hand placed details and environmental storytelling. Look at Skyrim/Cyberpunk vs. Starfield where most of the world is proc gen’d.
I get that you run this, and then edit it. But the generated villages just aren’t interesting in my opinion. It’s probably great for tencent’s market where you’re mass producing gacha style games, but I don’t think worlds made in this fashion will scratch the open world itch like the best open world games out there.
You know what's also good at storytelling? AI. Or else we wouldn't have writing awards being revoked when it comes out it was written by AI.
I wouldn't say we don't need artists or writers, but I think most people will be surprised how quickly fully AI driven pipelines will advance here.
I'd actually be curious to read an award winning ai book, all the ai books/interactive story generators I've tried have been... Ok at best... Do you know what novels have been written by ai and award winning?
Japanese isekai might not be your cup of tea, but one example is [1] which was way back in Jan, and models have improved a lot since then.
My sneaking suspicion is that there are a lot more cases than get reported, because authors are disincentivized to disclose AI usage, and readers by and large feel icky about reading AI work for reasons unrelated to quality.
I can totally see a weird future where a lot of what people like is AI generated, and everyone lies to each other about it. So it's hard to do research on this phenomenon.
[1] https://0115765.com/archives/165655
What award/book was this?
Interesting prompt to world pipeline. I wonder if there can be more style applied than just the "MMO cartoony" style? Like medieval Witcher or realistic Cyberpunk buildings?
I'm sure the answer is yes, even now there are pretty good model generators out there that can do realistic models. Experimental things like this just tend to go with cartoony styles since mistakes and lower quality (read: faster generation) can be chalked up to "lol", whereas realistic styles tend not to be forgiving with art mismatches and shading mistakes.
In modern polished games, the art team spends a lot of time tweaking things to make sure it looks great and coherent. Most people are not willing to have their agent run for a month to do this work. I don't know if it's been tried yet but I would be really interested to see the result.
Yeah any game ready asset has to be low poly. Even if it has 4k textures the polygon (vertex) count must be low to do anything meaningful (raycast, load many models).
Usually a combination of distance and frustum culling (breaking the model down into several that only load when you need to see it) with 4k or 8k textures can make a low poly game ready asset look very high quality
Looks impressive, haven't looked at the code yet, I'll at night.
There is no code. The GitHub repo is empty.
"LGTM - approve"
The assets look good, but for some reason the terrain looks too cartoonish to me.
This looks so cool. Nice work :D
I hate to be an HN bikeshedder but why is everything "Claw" recently?