Evolving Assembly Bots With LLMs: Core War Beats Zero-Shot, Says Akarsh Kumar
Machine Learning Street Talk · youtube · 2026-10-08
Machine Learning Street Talk released a short from its conversation with Akarsh Kumar about using LLMs to evolve small assembly programs that battle each other in the classic Core War game.
Per Kumar, zero-shotting the language works terribly, but combining evolutionary iteration with a verifier for selection turns out to be really effective — an LLM-as-mutator, verifier-as-selector loop for adversarial code generation.
More from coding & agent
- Hack: run coding agents in Grok's cloud computer to burn unused Code Plan credits — op7418 · 2026-10-08
- Grok bot surfaces $25,730 in live GitHub bounties, top one pays $10k — prasenx · 2026-10-08
- Talk: how to RL-train an agent running inside a harness you didn't write — SergioPaniego · 2026-10-08
- 8 open-source AI agent tools to know, from computer-use to browser automation — nikola_mr64990 · 2026-10-08
- Test quality follows module design: test larger units instead of banning AI tests — mattpocockuk · 2026-10-08
- Matt Pocock: AI test quality depends on codebase design, banning tests is wrong — mattpocockuk · 2026-10-08