Karpathy’s Pelican
Andrej Karpathy discusses the evolving capabilities of LLMs, specifically using Claude Opus 5 to generate complex 3D code from literary text. He notes that while LLMs show immense stamina for custom tasks, they currently lack the ability to natively audit their own visual or interactive output.
Why it matters
This highlights the shift from simple prompt-based AI tasks to complex, multi-hour generative workflows, while identifying critical gaps in AI self-correction and multimodal perception.
Andrej Karpathy on X: "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.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in