Public transcript

How Sam Altman Actually Uses Dots to Run OpenAI and His Life

By: Every
Length: 33 min
Language: English
Text from: the video's captions
Transcribed: 13 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

Dan Shipper interviews Sam Altman at OpenAI's DevDay about the day's launches, including Dots, Spaces, the Decisions API, and ChatGPT apps. Altman explains why he believes AI will drive a renaissance rather than an industrial revolution, with people remaining at the center of work. He shares personal examples of how his dot has changed his mornings and decision-making, and discusses the platform strategy, ultrafast inference, and the startup golden age.

Key points

Sam Altman says developers have never been able to do more, but the bar has never been higher because people want much more stuff.
He believes AI can enable a renaissance centered on people rather than an industrial revolution centered on machines.
His dot, Dotty, gave him back his mornings by filtering overnight issues and handling things with good judgment.
Astra was the key research breakthrough for Dots — the model had to get over the intelligence bar.
Spaces was built because existing software abstractions break down when multiple people and their AIs collaborate.
The Decisions API was inspired by Diogo and accelerated by the response to Jev, aiming for very fast, low-cost inference.
Altman says there are far more good ideas outside OpenAI than inside, which is why the platform and portable subscriptions matter.
He does all his prompting on ultrafast and expects 8x speed at a great price soon, then 100x faster.

Questions it answers

00:00DevDay and the future of work

What has the rise of coding models taught Sam Altman about the future of engineering and work?

Altman says developers have never been able to do more, but the bar has never been higher, because people simply want much more stuff. He was surprised on the upside by how much latent demand existed and how much people have remained the center of everything happening with AI.

Altman was surprised by how much latent demand for software exists in the world.
He believes people remain the center of everything happening with AI.

06:12Renaissance vs. industrial revolution

Why does Altman frame AI as a renaissance rather than an industrial revolution?

The renaissance was centered around people while the industrial revolution was centered around machines, and many people were turned into cogs. Altman is hopeful AI brings a wave of discovery, art, and company formation where people stay in control of their lives.

Altman rejects the idea that humans become pets of AI.
He hopes AI brings a wave of scientific discovery, art, and company formation.

08:49How Altman uses his dot

How does Sam Altman personally use Dots in his daily life?

His dot (named Dotty) filters overnight issues so he can keep his mornings for deep work or his kids. It has also proactively found a Slack screenshot he'd given up searching for, and warned Thibault about a demo bug minutes before the keynote.

The dot gave Altman back his creative mornings.
It found a screenshot in Slack that Codex couldn't locate.
Altman's dot is named Dotty.

17:03Spaces and live documents

What is Spaces and why did OpenAI build it?

Spaces bundles office-suite tools because existing software wasn't designed for people and their AIs to collaborate closely. Altman highlights 'live documents' that update automatically, and his personal auto-updating dashboard summarizing the business based on his Slack messages.

Current software abstractions break down for human-AI collaboration.
Altman's dashboard is a personal, auto-updating summary of the business.

20:27Decisions API and Jev

How does the Decisions API relate to Jev and the race between labs?

The Decisions API was inspired by Diogo and pushed further by the response to Jev, aiming for very fast, low-cost inference. Altman says OpenAI can't cover every new architecture but will try to deliver the main workhorse model at every point on the cost/speed curve.

Diogo, formerly of OpenAI, inspired the Decisions API.
OpenAI aims to be best on the Pareto curve of cost, speed, and modality.

22:36Apps, subscriptions, and startups

Why did OpenAI open up ChatGPT to apps and portable subscriptions?

Altman believes there are far more good ideas outside OpenAI than inside, and wants builders to connect with users. Letting people bring their ChatGPT subscription anywhere means users get full-power AI without developers worrying about model costs.

Users can bring their ChatGPT subscription into third-party apps.
Altman calls this a golden age for startups.
He would love to be a startup investor again.

26:50Building more with AI

How does AI change what Altman can build and how OpenAI operates?

Lowering friction means people build way more; Altman says DevDay's 22 launches wouldn't have been possible without AI help. Inspired by the conversation, he commits to shipping a plugin extension or maybe even a real OpenAI product himself.

Altman says OpenAI is not a gigantic company.
He commits on-air to shipping something himself.

30:05Ultrafast inference and democratization

What does ultrafast inference change and when will everyone get it?

Altman does all his prompting on ultrafast and says the immediate feedback loop makes a wild difference to iterative thinking. He says you can pay for it now, expects 8x at a great price in the not-distant future, then 100x faster, while a premium frontier will always exist.

Altman does all his prompting on ultrafast.
OpenAI's own chip will contribute to fast inference for everyone.
He expects 8x speed at a great price soon, then 100x.

Notes

DevDay and the Future of Work

Sam Altman says developers have never been able to do more, but the bar has never been higher — people simply want much more stuff.
He was surprised by how much latent demand existed in the world and how much people remain the center of everything happening with AI.
Jobs requiring little creativity and where we don't care about the person may go to AI, but Altman sees no shortage of new things to do.

Renaissance vs. Industrial Revolution

Altman frames AI as a potential renaissance (centered on people) rather than an industrial revolution (centered on machines, where people became cogs).
He rejects the "house cats" doomer vision and hopes for a wave of scientific discovery, art, and company formation.

Dots in Altman's Life

His dot, named Dotty, gave him back his mornings: it filters overnight issues so he keeps his deep-work or family time.
Thibault's dot warned him about a demo bug 5 minutes before the keynote.
Altman's dot found a Slack screenshot he'd given up searching for, and built 5-6 versions of a feature from his jotted notes.
Astra was the key research breakthrough — the model had to get over the intelligence bar.

Spaces and Live Documents

Spaces exists because current software abstractions break down when multiple people and their AIs collaborate.
Altman uses a personal, auto-updating "live" dashboard summarizing the business from his Slack messages.

Platform Strategy

The Decisions API was inspired by Diogo and accelerated by the response to Jev, pushing toward very fast, cheap inference.
ChatGPT apps and portable subscriptions reflect Altman's belief that there are more good ideas outside OpenAI than inside.
He calls this a golden age for startups and says he'd love to be a startup investor again.

Ultrafast and Democratization

Altman does all his prompting on ultrafast; the fast feedback loop changes iterative thinking dramatically.
He expects 8x speed at a great price in the not-distant future, then 100x, with OpenAI's own chip helping democratize fast inference.

Text from the video's own captions. Summary and notes written by AI from the transcript, so check anything important against the recording.