DeepMind Discovery team takes on the Busy Beaver (6) open problem with Lean-verified cryptids
TZahavy · x · 2026-10-21
The Google DeepMind Discovery team, in a project led by student researcher dsantosmarco, has spent the past year attacking one of the most beautiful open problems at the edge of computation: finding the 6th Busy Beaver number. Since halting is uncomputable, no universal algorithm can solve BB(6) — new mathematical techniques must be invented, making it an open-ended domain for mathematical creativity. Unlike AI math systems that solve human-posed theorems, this work targets discovering hidden problems. The attached image shows the first 10,000 steps of four Lean-verified potential BB(6) cryptids, each encoding a hard number theory problem, inspired by bbchallenge.org.
Related event: DeepMind Uses Gemini Agents to Discover and Verify 11 New Turing Machines(3 posts)→
More from Research
- Recursive's Flint: automated research builds open local coding model, lm_head at 4-bit — _rockt · 2026-10-22
- Tsinghua paper: letting agents manage their own KV cache cuts cost by 27.7% — dair_ai · 2026-10-22
- ETH researcher vibe-codes SuperMimic reproduction in 1 hour, LAFAN training in 1.5 GPUh — ChongZzZhang · 2026-10-22
- Effective receptive field in DiTs mirrors 2017 CNN finding, CS professor notes — CSProfKGD · 2026-10-22
- Flint cuts bike-demand validation RMSLE 30.8% in 3 hours of autonomous model iteration — cong_ml · 2026-10-22
- Flint runs auto-research on a laptop: 26% faster scheduling vs round-robin — cong_ml · 2026-10-22