Bitter lesson comes for multi-agent systems: RL-trained swarms beat hand-designed orchestration
emax · x · 2026-11-17
A debate on multi-agent system design. @barrnanas argues the bitter lesson is coming for multi-agent systems: instead of hand-designing agent roles, hierarchies, and communication protocols, let agents learn cooperation via RL — recent work shows swarms can self-organize, developing emergent coordination and division of labor that outperforms best-of-N. Reposter @AddiHD02 goes further: emergent bottom-up coordination (stigmergy) is the agent-native form of orchestration, and we should question whether models need post-training to collaborate at all, versus simply giving them the right environment at the harness level.
More from coding & agent
- Dev argues software factories are temporary; the real question is human-agent workspaces — heyneighbor · 2026-11-17
- Glass Health launches developer app with Agent Builder, Context Libraries and Glass 6.0 API — GlassHealthHQ · 2026-11-17
- Official Xcode plugin lands in Claude Code plugin marketplace: build, test, SwiftUI previews — ClaudeDevs · 2026-11-17
- Claude Code desktop adds iOS Simulator pane with iPhone Duo fold/unfold testing in public beta — ClaudeDevs · 2026-11-17
- Running 100+ coding agents: 'I need a super agent just to orchestrate my agents' — signulll · 2026-11-17
- Dev shows his agent-collaboration workspace designing a game character — trq212 · 2026-11-17