Run Fewer Agents: exe.dev argues task management is a band-aid, proposes fast models for human comms
sull · x · 2026-10-02
exe.dev's blog post "ETOOMANYTHINGS? Run Fewer Agents" argues that every "software factory" demo is really just task management: agents are slow, and the easy fix is spawning new agents to hide latency. But high concurrency wrecks human flow states. Their experiment: start tasks with a powerful model, then hand all human communication to a fast, tool-less model (Luna 6) with an intentionally short context, so users get real-time discussion and the bigger model resumes with rich feedback. The retweeter agrees that scrolling transcripts is a poor interaction mode at high tok/s.
Related event: exe.dev: Run fewer agents instead of building task dashboards(3 posts)→
More from coding & agent
- FT: US tech outages blamed on rush to fix bugs found by AI — nordicinst · 2026-11-19
- Claude Code overtakes Cursor and Copilot in job-post mentions ahead of Anthropic IPO — antgoldbloom · 2026-11-19
- Dev submits PR adding periodic checkpoints to GKE sandboxes so agents don't lose progress — yoimnotkesku · 2026-11-19
- AWS adds private CA support to Bedrock AgentCore Identity for connecting agents to self-hosted IdPs — AWS ML Blog · 2026-11-19
- Orchestrator's Cloud Agents graduate from beta to general availability — julianweisser · 2026-11-19
- Zero users, 33 pages shipped: dev crowdsources audit for AI-agent trust schema — TheHamzaShahbaz · 2026-11-19