Matt Shumer: Insane Multi-Agent Setups Don't Make You More Productive
daniel_mac8 · x · 2026-09-16
Matt Shumer argues that elaborate multi-agent setups people build online are fun but don't actually boost productivity — you spend more time building the system than doing real work. His advice: use one agent session that farms out tasks to others and acts as a manager; when something breaks, tell the main session and it adjusts the overall approach.
Daniel Mac8 replied that we're nowhere near the optimal interface for multi-agents, with plenty of work left to do.
Related event: Shumer: skip complex multi-agent systems, use one manager session(3 posts)→
More from coding & agent
- Dev builds 3D universe explorer with 20 galaxies, 767 planets using Codex GPT-6.1 Sol — OpenAIDevs · 2026-11-02
- ChatGPT, Claude and Perplexity's fetchers don't run JavaScript, leaving client-rendered sites invisible to AI — gaganghotra_ · 2026-11-02
- SGLang boosts multimodal inference >10% by replacing GPU memory bookkeeping with a dict lookup — charles_irl · 2026-11-02
- kivi: hold Fn, talk, and it types polished text anywhere on your Mac — itsOmSarraf_ · 2026-11-02
- A base LLM plus 11 lines of code is all you need for a decision model — helloiamleonie · 2026-11-02
- Fix codebase structure before asking AI agents to write tests, says engineer — arpit_bhayani · 2026-11-02