TBPN
TBPN

OpenAI DevDay | Sam Altman, Alexander Embiricos, Andrew Ambrosino, Peter Steinberger, Tibo Sottiaux, Kath Korevec, Casey Neistat

September 29, 2026

AI Summary

5 min read

OpenAI DevDay 2026: Personal Agents, Faster Models, and the Shape of What's Coming

OpenAI's fourth DevDay drew 2,500 attendees and brought roughly 20 product launches, but the headline number that set the tone was this: ChatGPT now serves more than 1.2 billion people every week, and over 35 million people build with it weekly using working codecs. The dominant theme was the arrival of "personal AI agents" — called Dots — alongside GPT-6.1 Sol, ChatGPT Spaces, and a cascade of infrastructure changes that reshape what developers and users can do with the platform. But the news outside the keynote was equally consequential: AMD acquired AI research lab World Labs for $8.2 billion, smart ring maker Oura delayed its IPO despite strong oversubscription, and Long Lake completed its $6.3 billion acquisition of AMX Global Business Travel.

Dots, Codex Cloud, and the Agentic Shift

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

What you'll learn

  • 1 (00:00) **Show Open & DevDay 2026 Headlines** - The hosts set the scene live from OpenAI DevDay 2026 and run through the major announcements.
  • 2 (01:56) **Oura IPO Delay & Market Uncertainty** - The team discusses the smart ring maker Oura delaying its IPO, questioning the stated reason of "market uncertainty."
  • 3 (05:51) **Acquisitions: Amex GBT & AMD/World Labs** - Two major deals are covered: the finalization of a travel giant merger and a massive AI chipmaker acquisition.
  • 4 (09:06) **Interview: Sam Altman on DevDay, Dots, and the Future** - Sam Altman joins for the third time to discuss the vision behind the new "dots" and the state of OpenAI.
  • 5 (15:05) **Altman on Proactivity, Culture, and Compute** - The conversation shifts to the risks of proactive AI, maintaining startup culture, and the compute roadmap.
  • 6 (20:15) **Altman on Forecasting and Alignment** - Altman discusses his forecasting accuracy and the critical, unsolved problem of AI alignment.
  • 7 (25:16) **Interview: Alexander Embiricos on Codex and Dots** - Alexander Embiricos discusses the evolution from coding agents to the new "dots" which represent a step-change in usability.

+ Full timestamped outline available in the app

Guests on this episode

Sam Altman Peter Steinberger

Show Notes

  • (09:00) - Sam Altman discusses how increasingly capable AI agents can simplify work and daily life while fostering an open ecosystem for developers. He also addresses OpenAI’s startup culture, expanding computing infrastructure, simplified model selection, rapid experimentation, and the unresolved scientific challenge of aligning superintelligent AI.
  • (25:17) - Alexander Embiricos, an OpenAI product leader, discusses “dots,” proactive AI agents designed to act as personal delegates across workplace tools. He explains how personal and specialist dots can autonomously handle tasks, support individuals and enterprises, and make advanced AI more accessible through flexible interfaces and integrations.
  • (41:54) - Andrew Ambrosino discusses leading OpenAI’s Codex desktop and Space teams and unifying Codex’s agentic capabilities with ChatGPT. He highlights cloud and local-machine integration, collaborative AI workspaces, dynamic interfaces, faster models, and OpenAI’s iterative, feedback-driven product development.
  • (54:27) - Peter Steinberger discusses founding OpenClaw and pioneering personal AI agents that can proactively operate computers, use tools, and coordinate complex tasks. He explores agent-driven software development, changing online services, and a future where AI reshapes operating systems, commerce, and human workflows.
  • (01:16:22) - Tibo Sottiaux discusses his work leading AI agent development at OpenAI, including Dots, Codex, and advances in computer use. He explains how persistent agents can learn users’ preferences, securely perform specialized organizational tasks, share long-term memory, and automatically balance speed, cost, and reasoning effort.
  • (01:33:41) - Kath Korevec discusses her work as an OpenAI product manager for ChatGPT Sites, which reached 10 million sites within three months of launch. She highlights personal and professional uses, collaborative features within Spaces, new integrations and plugins, and plans to expand access to free users.
  • (01:41:50) - Casey Neistat discusses his perspective as a filmmaker, YouTube pioneer, and ModRetro executive on retro technology, AI-assisted creativity, and the value of simple, purpose-built devices. He argues that AI will democratize game and video creation while human taste, expertise, and curation remain essential amid a growing flood of content.


TBPN is made possible by:

Ramp - https://ramp.com

Public - https://public.com

Cisco - https://www.cisco.com

Console - https://www.console.com

CrowdStrike - https://www.crowdstrike.com

Figma - https://www.figma.com

MongoDB - https://www.mongodb.com

NYSE - https://www.nyse.com

Railway - https://railway.com

Shopify - https://www.shopify.com

Codex - http://openAI.com/codex


Follow TBPN: 

https://TBPN.com

https://x.com/tbpn

https://open.spotify.com/show/2L

TBPN

More from this podcast

TBPN →