Greg Brockman (Part 1) - Tetragrammaton with Rick Rubin

Transcript

Podcast Host / Narrator (0:02)

Tetragrammaton.

Greg Brockman (0:23)

One thing that's really changed over the course of 2025 is, was people started to use ChatGPT for much more personal, very intimate applications. For example, so my wife has complex medical conditions, including hypermobile Ehlers Danlos syndrome, which took many years for us to get diagnosed. And as we put those symptoms into ChatGPT, it would be able to figure out pretty immediately. But the thing is that every doctor has their own specialty or rather than one doctor who can see across everything, and she uses ChatGPT to manage her health all the time.

Interviewer / Podcast Host (1:00)

Great.

Greg Brockman (1:00)

For me, the funny thing is I'm actually a late adopter of our own technologies usually, and I usually test them and I stress test them in all sorts of ways. And it's funny, for early versions of our models, I would usually like try to break some of their filters. And so I would swear at them and yell at them. And my wife is always like, he's just kidding, you know, telling me to be nice to the AI. And I think that for me, I actually have been someone who's almost very set in my ways. And the first time this has changed is very recently with Codex and really starting in December, like I've been a curmudgeon, I've been, I've got my way of doing things, I use my terminal, I use my emacs, like all these like tools that I grew up with and I've just abandon all of that now.

Interviewer / Podcast Host (1:45)

Wow, just using Codex, that's revolutionary.

Greg Brockman (1:49)

It really is.

Interviewer / Podcast Host (1:49)

You sound like you were set in your ways.

Greg Brockman (1:51)

I really was, yes.

Interviewer / Podcast Host (1:53)

What changes with each new model?

Greg Brockman (1:56)

Everything. And the way to think about it is that we, from the outside perception is, oh, you're just scaling up the models, you're just doing this kind of dumb thing on the inside. Every single part of the process, we are always up leveling. Like the thing that works in machine learning and that machine learning rewards is attention to detail. And so you really want to make sure that all of the scaling is right. You want to make sure that the systems, for example, GPUs failing, that happens. And so how do you detect if you have a run of 100,000 GPUs, how do you detect which GPU is the broken one? It's not easy, right? You can't just be like that one. Right. So you need this almost. There's a physical process to it, there's the software process to it, there's understanding if your data is any good and that there's so much reward to just actually Looking at the data to understand what's in there and to make sure that you format it correctly and it's tokenized properly. And so there's just every single part of the input we are constantly improving and we pay attention a lot to the output too. We're trying to see how does improvement here connect to an eval shifting. And one observation I have across many years of OpenAI is that if we have some signs of life, some application that kind of works right now, one year from now, you should expect it to be excellent. And so we are on this exponential and able to make these like very, very sophisticated improvements over time.

Summary

Tetragrammaton with Rick Rubin — Greg Brockman (Part 1)
Date: February 25, 2026

Overview

In this engaging and deeply personal conversation, Rick Rubin sits down with Greg Brockman—President and Co-founder of OpenAI—for an intimate exploration of AI's incredible growth, the human drama inside OpenAI, technical realities, and the emotional and philosophical stakes of building artificial general intelligence. Brockman shares anecdotes from the internal power struggle that briefly ousted Sam Altman as CEO, reflects on OpenAI’s rapid growth, traces decisions with major partners like Microsoft, and speculates on where AI—and OpenAI itself—might head next.

Key Discussion Points & Insights

1. AI in Daily Life & Personal Reflections (00:23–03:19)

Personal Use of AI: Brockman describes how his wife, who has complex medical conditions, uses ChatGPT to manage her health, diagnosing quickly what took doctors years.
Shifting Perspectives: Greg admits to being a "late adopter" of their own products, only recently abandoning his old-school programmer tools for newer AI-fueled products like Codex.
- Quote: “I've got my way of doing things, I use my terminal, I use my emacs, like all these tools that I grew up with and I've just abandoned all of that now.” (01:00)
The Perception of Progress: He emphasizes that scaling AI is not just ‘dumb scaling’ but involves constant attention to detail, hardware, and data.

2. Inside OpenAI: Culture, Crisis & Leadership (03:19–19:52)

On Sam Altman

Sam Altman as Co-founder: Brockman addresses Sam’s polarizing reputation, describing him as “a very good person... it's very much no good deed goes unpunished.” (03:25)
- Quote: "Every accusation is a confession. People project onto him things that they see in themselves they're insecure about...” (03:25)

The Firing Incident & Leadership Turmoil (05:13–18:17)

Recap of Sam’s Firing: Brockman recounts the chaotic boardroom drama that led to Altman’s ouster and his own resignation from the OpenAI board—events he describes as shocking yet, given simmering conflicts, almost inevitable.
Employee Rebellion: Employees rallied, nearly all signed a petition demanding leadership’s return: “So many people were editing that Google Doc for the petition that actually crashed Google Docs.” (16:35)
Microsoft’s Role: During the crisis, Microsoft offered to ‘take everyone’ if necessary, expressing commitment to the mission and team.
Forgiveness & Resolve: Ilya Sutskever’s public regret was an emotional turning point for Brockman, who talks about finding forgiveness and renewed focus.
- Quote: “I felt forgiveness. I felt like, okay, beautiful.” (17:35)

Lessons Learned

The Importance of Hard Conversations: Post-crisis, Brockman says the OpenAI team now values early, direct resolution of internal conflict. “Have the hard conversation.” (13:14)
Mortality & Humility: The ordeal instilled a sense that nothing is preordained and that internal human conflict, not technology, is the largest threat.
- Quote: “The thing that is most likely to get in the way of success is tripping over ourselves.” (19:10)

3. Origins, Structure, & Major Partnerships (23:37–35:31)

On Elon Musk & OpenAI’s For-Profit Pivot

Disagreements with Elon Musk: Brockman recounts intense negotiations. Musk wanted absolute control; the team demurred, fearing AGI concentrated in a single person’s hands.
- Quote: “Imagine you actually build an AGI and there’s one person who has absolute control over it... Do you feel good?” (24:49)
Alternative Funding Paths: Tales of seeking capital, including nearly doing a crypto ICO, and the harsh reality of Musk’s skepticism. Musk told them OpenAI had “0% chance of success relative to Google without a dramatic change in resources and execution.” (29:01)

Microsoft Partnership

Early Collaborations: Started with compute donations for OpenAI’s DOTA project.
Microsoft Investments: Ultimately provided $1B, then $2B and $10B more—enabling OpenAI to design massive supercomputers necessary for training cutting-edge models.
- Quote: “No one else was putting in that quantum capital at the time.” (33:32)

4. Philosophy, Product Development, and AI’s Capabilities (35:34–44:40)

Is ChatGPT Religious? Not inherently, says Brockman: “You should be able to shape it to your own preferences... ChatGPT is not an entity but... a plurality... that can be shaped in different form.” (35:36)
Emotions/Humanity in AI: Currently, AI lacks a proper theory of emotion, but it's a philosophical and technical issue OpenAI knows it must address as models grow smarter.
Coding & Leadership: Brockman remains hands-on, believing in “leading from the trenches” and learning the right direction by direct technical work—even as his responsibilities have grown.
- Quote: “In this field, because things change so quickly... you need to really feel it.” (40:02)
On AI Breakthroughs: Recent months have seen explosive improvements in software capabilities, outpacing expectations for areas like debugging, but AI is still lacking in high-level human judgement and negotiation.
- Quote: “Can you have models that help negotiate people have differences? ... I don’t really see the models yet adding value.” (38:48)

5. AI, Reasoning, and the Road Ahead (53:49–54:48, 56:42–57:51)

The Exponential Curve: From now until 2027, Brockman predicts transformative “agent” technology for knowledge work, scientific research, and a generative explosion in new companies fueled by AI.
- Quote: “2025, 2026, 2027 are going to be these transformative moments... you can really see it this year. We’re going to have agents for knowledge work for every single function.” (53:53)
Reasoning AI: Brockman describes the move from immediate-answer AIs to ones capable of expending compute to reason, revise, and arrive at better answers—mirroring human assistants.
The Importance of Diversity in AI: Rather than a monolithic “aligned” entity, Brockman values a resilient, diverse ecosystem of AI agents and organizations to ensure safety and flexibility for humanity.

6. The State & Shape of OpenAI (49:56–51:16, 60:22–67:08)

Organization Size & Roles: OpenAI now has around 5,000 employees—1,000–1,500 in research, others in engineering, product, and operations.
Not Yet Profitable: “We are not a profitable company. We spend a ton on compute.” (34:26, 60:28)
Compute as a Bottleneck: The scarcity of GPUs and physical computing resources is now the limiting factor in growth; Brockman believes “compute as a basic human right” will become a reality.
Resilience & AI Safety: The goal is a diverse, decentralized AI ecosystem, resilient to monoculture and equipped to address safety risks.

7. Reflections on Learning and Leading (54:48–67:08)

Constant Change: Brockman’s role and focus shift dramatically every few months, requiring continual learning and adaptation.
- Quote: “At every single stage... I look back at last year me, I look back at five year ago me, and I'm like, I knew nothing, like really nothing.” (51:16)
Comparison to Stripe: OpenAI functions less like a traditional machine-building company (Stripe) and more like a movie studio—constantly needing a new hit, with research and reinvention as foundational.
Would He Do Anything Differently? Brockman expresses pride in how OpenAI has steered its ship, making “principled, reasoned decisions at every step,” and sees their hybrid nonprofit/for-profit structure as essential and forward-thinking. (66:00)

8. Technical, Linguistic, and Societal Implications (70:19–76:34)

Unsupervised Sentiment Neuron: OpenAI’s 2017 paper was a pivotal moment, proving that language models could internalize and recognize complex sentiment—heralding a dramatic leap forward.
Linguistics & AI: Modern models arguably “capture the true underlying rules” of language, with capabilities in understanding sarcasm, double entendre, poetry, but limited in crafting jokes (for now).
Distributional Knowledge: Brockman explains how AI generates a distribution over possible answers, not one deterministic reply, fueling creativity and breadth.

9. Personal Journey & Upbringing (77:13–89:45)

Early Education: Raised in rural North Dakota, accelerated quickly through math and science—helped by supportive parents (both doctors) and flexible schools.
MIT vs. Harvard: Ultimately transferred from Harvard to MIT, seeking deeper technical challenges and a more hacker-centric culture: “Looking back, I’ve realized that many of my life decisions are made by me dreaming of the upside...” (78:32)
Acting & Improv: Early love of acting and improv fueled a creative approach to life and technology; sees a close link between improv (“yes, and...”) and the way AI models adapt and build on context in real time.

Notable Quotes

“Have the hard conversation.” (13:14 – What he learned after OpenAI’s near-collapse)
"Imagine you actually build an AGI and there’s one person who has absolute control over it... Do you feel good?” (24:49 – On not granting Elon Musk total control)
“We want something to generate something new... deeply understand whatever is put in front of them. And if you think about how you as a human learn, it's not that different.” (73:39)

Important Timestamps

00:23–03:19: Brockman on personal adoption of ChatGPT and shifting from old tools to Codex
05:13–18:17: The firing of Sam Altman, Brockman’s resignation, employee revolt, and return
23:37–26:20: Musk’s control demands, failed negotiations, and eventual split
32:51–35:31: Microsoft partnership and scale-up
53:49–54:48: Five- and ten-year projections for AI and OpenAI
60:22–63:12: OpenAI finances, compute limitations, and the compute arms race
70:19–73:39: Sentiment neuron, linguistics, and emergent AI abilities
77:13–89:45: Brockman’s upbringing, education, and path to MIT

Memorable Moments

The Google Doc Petition Crashing Google Docs: (16:35) Symbolizing unprecedented employee unity during the OpenAI crisis.
Ilya's Public Regret & Forgiveness: (17:00–17:35) A deeply human moment in the midst of technical and corporate drama.
Improvised Gingerbread Man: Greg’s acting career peaked (89:45) after being too small to be a tin soldier, teaching him about adaptation and seizing unexpected opportunities.
MIT's Hacker Culture: Lights with individual IPs and a proprietary chat system illustrate Brockman’s shape-shifting technical journey. (79:28–81:29)

Tone & Style

The conversation is at once candid, technical, and emotional—alternating between big-picture philosophical questions and deeply personal anecdotes about leadership, burnout, learning, and resilience. Rick Rubin keeps the tone curious, open, and contemplative, while Brockman is honest, methodical, and sometimes vulnerable.

Conclusion

This episode offers an unprecedented window into the challenges, victories, and philosophical struggles inside OpenAI, as seen by one of its co-founders. It fuses technical depth with human vulnerability and sense of mission—a must-listen for anyone fascinated by the intersection of technology and humanity.

End of Part 1. (Part 2 is teased for release later in the week.)