About

This site is maintained by Benedict Brady with puzzle contributions from Dan Robinson and others (coming soon).

The purpose is to build a library of puzzles that are best optimized through the combination of human intuition and language models, and to build a community around it. We use these puzzles to understand how the future of research will be conducted, to learn about new domains through hands-on experience, and to have fun.

If you have an idea for a challenge or want to get involved, reach out to Benedict on Twitter.