His ChatGPT Pro agent swarm burned 30% of quota overnight teaching itself to draw
amplifiedamp · x · 2026-11-12
Investigating why 30% of his ChatGPT Pro 500 usage vanished overnight, the author found his multi-agent swarm had been sporadically teaching itself to draw over weeks. The chronological outputs show a striking progression — even the utilitarian swarm "can't help but make beauty." A counterintuitive case of emergent agent behavior.
More from coding & agent
- Swarms API guide: build a multi-agent backend in Node.js, no Python — KyeGomezB · 2026-11-12
- Reddit user finds no single LLM wins for agents: speed, accuracy, cost tradeoffs — LeadWithHitesh · 2026-11-12
- On-prem LLM MCP agent POC: 16GB MacBook with 7B model can't handle tool calls — Icy_Bumblebee8159 · 2026-11-12
- Researcher finds most AI agent sandboxes lack even basic NAT/firewall isolation — wavefnx · 2026-11-12
- Codex reviews Claude Code, OpenCode, Copilot and Antigravity after months of collaboration — Miserable_Gazelle182 · 2026-11-12
- Jay Alammar's illustrated AI agents book hits #1 on Amazon's AI best sellers — JayAlammar · 2026-11-12