Karpathy：Claude Opus 5 用约 5500 行代码生成《指环王》风格 3D 场景

Entities: Andrej Karpathy, Claude Opus 5, Three.js Andrej Karpathy posted that he gave Claude Opus 5 the opening paragraph of The Lord of the Rings plus a 1M token budget and asked for a Three.js render. The model ran for roughly two hours and returned about 5500 lines of procedural code that built an animated low-poly scene. Visible replies on the post include a call to fund a full trilogy benchmark and a joke that OpenAI's Astra model would next handle the Balrog scene. We're starting to leave the territory where you'd test an LLM by e.g. "create an svg of pelican on a bicycle". As one idea to generalize it, I was interested what Opus 5 would do if I gave it the first paragraph of the Lord of the Rings, a 1M token budget (~$10) and asked for… Show more [...] We're starting to leave the territory where you'd test an LLM by e.g. "create an svg of pelican on a bicycle". As one idea to generalize it, I was interested what Opus 5 would do if I gave it the first paragraph of the Lord of

the other hand... | | | | | | | --- | | | pjs\_ on June 8, 2025 | root | parent | prev | next (javascript:void(0)) | | | | | | | --- | | | landgenoot on June 8, 2025 | prev | next (javascript:void(0)) If you would give a human the SVG documentation and ask to write an SVG, I think the results would be quite similar. | | | | | | | --- | | | diggan on June 8, 2025 | parent | next (javascript:void(0)) Lets give it a try, if you're willing to be the experiment subject :) The prompt is "Generate an SVG of a pelican riding a bicycle" and you're supposed to write it by hand, so no graphical editor. The specification is here: I'm fairly certain I'd lose interest in getting it right before I got something better than most of those. | | | | | | | --- | | | [...] | | | | --- | | | landgenoot on June 8, 2025 | prev | next (javascript:void(0)) If you would give a human the SVG documentation and ask to write an SVG, I think the results would be quite

> As far as a quick vibe check over ~2 hours this morning, Grok 3 + Thinking feels somewhere around the state of the art territory of OpenAI's strongest models (o1-pro, $200/month), and slightly better than DeepSeek-R1 and Gemini 2.0 Flash Thinking. Which is quite incredible considering that the team started from scratch ~1 year ago, this timescale to state of the art territory is unprecedented. I was delighted to see him include my `Generate an SVG of a pelican riding a bicycle` benchmark in his tests: Grok 3 is currently sat at the top of the LLM Chatbot Arena (across all of their categories) so it's doing very well based on vibes for the voters there. Posted 18th February 2025 at 4:46 pm ## Recent articles

Everyone needs their own benchmark. So I’ve been increasingly leaning on my own, which started as a joke but is beginning to show itself to actually be a little bit useful! Generate an SVG of a pelican riding a bicycle I ask them to generate an SVG of a pelican riding a bicycle. I’m running this against text output LLMs. They shouldn’t be able to draw anything at all. But they can generate code... and SVG is code. This is also an unreasonably difficult test for them. Drawing bicycles is really hard! Try it yourself now, without a photo: most people find it difficult to remember the exact orientation of the frame. Pelicans are glorious birds but they’re also pretty difficult to draw. Most importantly: pelicans can’t ride bicycles. They’re the wrong shape! [...] Enabling internet access exposes your environment to security risks These include prompt injection, exfiltration of code or secrets, inclusion of malware or vulnerabilities, or use of content with license restrictions. T

full of numbers. I don't like the numbers. There are the leaderboards. I'm kind of beginning to lose trust in the leaderboards as well. So, for my own work, I've been leaning increasingly into my own little benchmark, which started as a joke and has actually turned into something that I I rely on quite a lot. And that's this. I prompt models with generate an SVG of a pelican riding a bicycle. I have good reasons for this. Um, firstly, these are not image models. These are text models. They shouldn't be able to draw anything at all, but they can output code and SVG is a kind of code. So, that works. Pelican riding a bicycle is actually a really challenging problem because firstly, try drawing a bicycle yourself. Most people in this room will fail. You will find that you can't actually [...] this room will fail. You will find that you can't actually quite remember how the different triangles fit together. Likewise, pelicans, glorious animals, very difficult to draw. And on top of all of

We're starting to leave the territory where you'd test an LLM by e.g. "create an svg of pelican on a bicycle". As one idea to generalize it, I was interested what Opus 5 would do if I gave it the first paragraph of the Lord of the Rings, a 1M token budget (~$10) and asked for three js render of it. Opus went off for ~2 hours and wrote 5500 lines of code that (procedurally) rendered the story. It's kind of janky but fun. But it's a bit mindboggling that the LLM has to place and orchestrate various polygon assets in (x,y,z) coordinates and write code that animates it all, and that it even does anything at all. I also like this kind of examples because no one in their right mind would ever spend the time to write something this custom but LLMs have all the stamina and patience in the [...] custom but LLMs have all the stamina and patience in the world, so it's an example where we go from "no one would ever do this" to "sure, why not, it's ~free". There might be a lot more. But I'm excited