ResearchOctober 01, 2026
15
🌱
Stratego AI Beats World Champion for $8K
An AI called Ataraxos defeats Stratego's greatest player, ending one of the last human strongholds in board games.
#Stratego#DeepMind#Game AI#Reinforcement Learning#Nature
🔥 What happened
A team from CMU, NYU, Stanford, and MIT built Ataraxos, the first AI to reach superhuman play in Stratego. In a 20-game series it crushed Pim Niemeijer, the most decorated Stratego player ever, 15-1 with four draws.
💡 Why it matters
Training cost under $8,000 on 16 H100 GPUs — roughly 1/500th of DeepNash's estimated $3–4.5M budget. Stratego, with over 10^33 possible setups, was one of the last board games where humans still ruled.
⚡ Our take
A university budget just beat what DeepMind's millions couldn't — and DeepNash's code doesn't even run anymore. Sample efficiency is the real story here, not raw compute.
The title, summary and analysis of this item were produced automatically by an AI system and have not been editorially reviewed. They may contain errors, bias or omissions — when in doubt, read the linked original source.
