Dev runs 10-12 parallel coding sessions; says decision-making becomes the bottleneck
robleclerc · x · 2026-11-01
Rob Leclerc shares his AI coding workflow: he manages 10-12 parallel coding sessions at a time — necessary because per-session latency is high, but it caps total usage. Smarter, more agentic models will push that ceiling far higher.
At peak productivity, he argues, you'll feel like Elon Musk: focusing only on the essential bottlenecks and decisions while delegating to tens of thousands of agents. The limiting factors will be your ability to make fast quality decisions, know when to go deep on what's underneath, and how much feedback AI can gather on its own versus what still requires human interface with reality.
Quoted context: token prices are down 58% since July while GPU rents rose 17-106% — cheaper intelligence drives usage growth that outpaces efficiency gains, giving pricing power to GPU owners.
More from coding & agent
- 747 AI-agent ideas reveal a scaling law for idea descent: +17% per doubling — DimitrisPapail · 2026-11-01
- Magnific Move + Claude Opus: Halloween game animated in one chat — petewoodbridge · 2026-11-01
- AI-Driven Frame Rate Optimization Cuts Render Time by 9%-15% Across Six Scenes — AIandDesign · 2026-11-01
- Parcha founder: You may not need an agent after all — decision models vs workflows — blaizedsouza · 2026-11-01
- Claude Code agents burned 9.45M tokens a day reading their own notes — WWEtrump · 2026-11-01
- Open-source agent filters Reddit feed with cheap LLM, under 1 cent per 100 posts — Ill_Fisherman8352 · 2026-11-01