Karpathy’s Pelican: A New Frontier in AI

• Opus 5 successfully created a playable pinball game
• LLMs struggle with physical world understanding
• New benchmarks are needed to measure progress

The Buzz Score

The Internet’s Verdict: 70% Hyped, 30% Skeptical

Expert Insights

Experts are weighing in on the capabilities of Karpathy’s Pelican.

A simple prompt that still stumps frontier LLMs most of the time is “create a pinball game”. They’ll put all the right pieces there but then fail to arrange them such that the game is truly playable.

Models are moving beyond image generation to a new kind of benchmark that better exposes understanding of the physical world.

A lot of people are posting here about how bad the end product is, but that is kind of the point. Models have moved beyond generating images to a new kind of benchmark that better exposes understanding of the physical world, and we can use benchmarks like this to measure future progress.

Real-World Applications

Developers are using LLMs to build complex animations and scenes.
For example, a 3D animation of the Back to the Future delorean Time Machine was built using an LLM.

Focus Keyword: Karpathy Pelican