'Jev' plays Pac-Man steered by reasoning model Astra in early dual-agent demo
daniel_mac8 · x · 2026-09-17
danielmac8 demos a fun two-model setup: Astra strategizes while 'Jev' carries out the plan in milliseconds to play Pac-Man. Early and hard to fully parse, he says, but combining 'Jev' with larger reasoning models clearly has potential.
Related event: Dual-AI Demo Splits Planning and Execution to Beat Pac-Man(2 posts)→
More from coding & agent
- A practical prompt recipe for AI code review: solve it yourself first, then compare — dotey · 2026-09-17
- Open-Sourced Qwen-2.5-1B-RLCD: 5x Faster On-Device Inference for JSON Workloads — max_paperclips · 2026-09-17
- RAGForge released: prototype full RAG pipelines in your browser, embeddings and LLM included — IanArawjo · 2026-09-17
- pi-crew brings non-blocking parallel subagent orchestration to the pi terminal AI — solyarisoftware · 2026-09-17
- ChainForge v0.3.7 adds RAGForge pipelines, in-browser LLMs and PDF uploads — IanArawjo · 2026-09-17
- Dev wires Claude Code to Grokbot via HQ DMs to build better integration — jacob_posel · 2026-09-17