Tibo on Lenny's Podcast: give your Dot goals and feedback, not hand-built loops — and multi-Dots are coming

#What happened

Thibault Sottiaux (Tibo), OpenAI's Head of ChatGPT, appeared on Lenny's Podcast (episode: "OpenAI's Head of ChatGPT: We're entering a new era of AI (again)") discussing Dots, the always-on agents OpenAI launched at DevDay on September 29. In the interview, he said people got excited about setting up and fiddling with their own loops, "but I don't think this is the way that it's going to work" — instead, a Dot "works 24-7, understands your goals, understands your preferences, learns from feedback." He also revealed that users will "very quickly" be able to add a second, third, fourth dot — "as many as you need to create your virtual team."

#More roadmap signals

• Dots capabilities will eventually ship straight into ChatGPT itself, raising the floor for what Tibo called "our 1.2 billion users."
• Dots have no model picker and no configuration: "The only thing that you have to configure is which channels you want to talk to."
• Codex and ChatGPT are merging their split chat/work surfaces: "we're reducing that complexity."
• On agent scale, Tibo's anecdote: his own Dot warned him about a production outage five minutes before a launch demo.

#Why it matters

Tibo's framing redefines the job of a Dot: instead of users hand-building loops, Dots should learn goals and preferences from feedback. With multi-Dots arriving "very quickly" and Dots capabilities eventually shipping straight into ChatGPT — raising the floor for "our 1.2 billion users" — the feedback-driven model is the path to scale.

#Sources

• Episode transcript page
• Episode page

Verified October 5, 2026.