Dev Interrupted Scaffolding is coping not scaling, and other lessons from Codex | OpenAI’s Thibault Sottiaux
18 snips
Jan 27, 2026 
Thibault Sottiaux, an OpenAI Codex engineer who builds agentic coding systems, explains why removing scaffolding leads to true agent autonomy. He discusses agent-first design, simple scalable primitives over clever harnesses, open-sourcing Codex tooling, multi-agent scaling, portability trade-offs, and how agents reshape developer workflows and careers.
AI Snips
Chapters
Transcript
Episode notes
Simplicity Prevents Capability Overhang
- Simplicity in the harness reduces constraints that create capability overhangs as models improve.
- Sottiaux recommends choosing primitives that scale with model capabilities to keep the system extensible.
Open Source To Demystify Agents
- Open sourcing the Codex repo aimed to demystify agents and show how simple primitives can yield strong performance.
- Sottiaux wanted to learn how open source itself would change code generation and to invite community tinkering.
Community Course Correction
- Early community management missteps occurred when the repo accepted too many contributions and lost control.
- The team migrated core code to Rust and rebuilt partnerships with key contributors to regain direction.