Jun 21, 2025 · 42m · y-combinator
Sam Altman: The Future of OpenAI, ChatGPT's Origins, and Building AI Hardware · Y Combinator
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
At Y Combinator's AI Startup School, OpenAI CEO Sam Altman and YC President Garry Tan discuss the origins of OpenAI, the rapid evolution of AI capabilities, strategic advice for early-stage founders, and the societal impact of superintelligence. Altman highlights how reasoning agents, compute energy bottlenecks, and technological optimism will shape the future of tech entrepreneurship and human capability.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The partners as informed peer 2.9 Guest teaching 2.9 Guest disagreement 0.9 The partners pushing back 0.1
speaking balance: gold is the partners, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 37:04 Dismissing the degrowth philosophy
Altman sharply rejects the degrowth mindset, contrasting startup optimism with European degrowth conferences and stating firmly that technologists are right and they are wrong.
Hardest push from the partners ▶ 30:56 Clarifying hiring criteriaTan questions what a strong track record actually means, pressing on whether founders should hire experienced big-name administrators or technical builders.
Biggest teaching moment ▶ 3:55 Debunking the overnight scale mythAltman reframes Tan's question about starting huge companies, educating the audience that OpenAI began as an uncertain research project of eight people in a room.
The partners hold their own ▶ 22:28 Formulating the JIT software paradigmTan articulates an architectural vision of business software where LLM agents dynamically generate interfaces directly over database APIs, which Altman validates.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The partners as informed peer | Guest teaching | Guest disagreement | The partners pushing back | Why |
|---|---|---|---|---|---|---|
| The Founding Decisions Behind OpenAI | 2 | 3 | 1 | 0 | Gary introduces early pivotal decisions and scaling laws. Sam contextualizes how improbable OpenAI seemed initially compared to DeepMind and how primitive early prototypes were. | |
| Gathering World-Class Talent Around AGI | 2 | 4 | 1 | 0 | Gary asks if founders should try starting massive scale projects immediately. Sam dispels the myth, noting OpenAI started with eight people in a room and quotes Vinod Khosla on zero million versus zero billion startups. | |
| Model Capabilities, Cost Reductions, and Open Source | 3 | 3 | 1 | 0 | Gary brings up user memory and emergent model capabilities. Sam elaborates on the product overhang, dramatic drop in inference costs, and future open-source model releases. | |
| Data Integrations, Model Context Protocol, and Hybrid Infrastructure | 3 | 3 | 0 | 0 | Gary notes internal YC agent usage and rumors about MCP integrations. Sam discusses ChatGPT evolving into an operating system and the computational challenge of scaling infrastructure. | |
| Rebuilding American Industrial and Manufacturing Capacity | 4 | 4 | 2 | 1 | Gary raises defensibility and industrial manufacturing bottlenecks. Sam gives direct advice warning startups not to copy ChatGPT's core assistant features and to avoid crowded startup ideas. | |
| Maintaining Conviction and Enduring Pushback | 3 | 2 | 1 | 0 | Gary cites Peter Thiel's contrarian philosophy. Sam reflects openly on the emotional difficulty of facing severe pushback, recalling a discouraging email from Elon Musk. | |
| AI Agents and Interfaces That Melt Away | 3 | 3 | 1 | 0 | Gary asks about workflows and hiring Jony Ive. Sam explains his vision for human-computer interaction where stressful phone interfaces melt away into trusted background agents. | |
| Just-In-Time Software and Startup Speed Advantages | 5 | 3 | 0 | 0 | Gary outlines a technical architectural thesis of JIT software replacing traditional SaaS. Sam strongly validates the thesis and explains why rapid clock cycles favor startups over incumbents. | |
| The Age of Intelligence and Individual Empowerment | 3 | 2 | 0 | 0 | Gary cites business strategy literature while Sam reflects on historical technological scaffolding that allows small teams with high agency to achieve massive leverage. | |
| Living on the Leading Edge of Technological Change | 3 | 3 | 1 | 1 | Gary asks how founders should hire and whether to look for eminent administrators. Sam rejects hiring polished administrators early on, advocating hiring for trajectory and demonstrated output. | |
| The Intensity and Context-Switching of the OpenAI CEO Role | 3 | 3 | 1 | 0 | Gary touches on the energy-intelligence relationship and historical correlations. Sam admits he originally viewed energy and AI as orthogonal before realizing energy would constrain compute. | |
| Technological Optimism Versus the Degrowth Mindset | 2 | 2 | 2 | 0 | Gary and Sam discuss technological optimism versus degrowth, with Sam drawing a comparison to the invention of the transistor accelerating societal welfare. | |
| Flashback to Y Combinator 2005 and Paul Graham | 2 | 2 | 1 | 0 | Gary and Sam close by reminiscing on YC's 2005 batch and Paul Graham, with Sam sharing lessons on long-term conviction, resilience, and parallels to parenthood. |
Statements from this episode (33)
Altman: Founding team came very close to not starting OpenAI
“Like, there, we got very close to not starting OpenAI.”
Altman: OpenAI started with no product ideas, revenue, or revenue expectations
“We were trying to, like, play video games, and we had this little robotic hand that could sort of barely do a Rubik's Cube, and we had no ideas for products, no revenue, no really idea that we were ever going to have revenue, and we were, like, sitting around …”
Altman: Unique Missions Give Startups a Strong Tailwind in Recruiting Top Talent
“If you are doing the same thing as everyone else, it is very hard to concentrate talent, and it's very hard to get people to, like, really believe in a mission, and if you're doing, like, a one-of-one thing you have a really nice tailwind there.”
Altman: OpenAI's o3 model cost dropped by 5x in one week
“And also, like last week, O-three cost five times as much as it did this week, and that's gonna keep going.”
Altman: OpenAI has an open-source model releasing soon
“We have an open source model coming out soon.”
Altman: OpenAI's open-source model will run locally and surprise people
“I think it will be, like, much better than you're hoping for, and the ability to, like, use it to run incredibly powerful models locally is gonna, like, really, really surprise people on what's possible.”
Altman predicts AI will shift from turn-based chat to proactive persistent companions
“You will have this, like, entity that gets to know you, that connects to all your stuff, and that is, like, proactively helping you. It'll, it won't just be like you send a message and it sends you one back, but it'll, like, be running all the time. It'll be l…”
Altman: AI companion value lies in background persistence, not hardware
“I think the key of her is not the little piece of hardware. It's that this thing got to a point where it could run in the background and feel like a sort of AI companion.”
Altman: OpenAI will soon run world's largest, most expensive infrastructure
“I think we will run the, like, largest and most expensive piece of infrastructure in the world pretty soon.”
Altman: ChatGPT is now fifth largest website in the world
“ChatGPT.com didn't exist two and a half years ago to, like, the fifth biggest website in the world.”
Altman: ChatGPT will become third biggest website, possibly first
“It'll be the third at some point, hopefully someday the first if our current growth rates continue”
Altman: GPT-5 will lack complete reasoning and real-time video integration
“We won't get all the way here with GPT-V, but eventually we do want one integrated model that can, like, reason really hard when it needs to and generate, like, real-time video when it needs to do that.”
Altman predicts robots will perform super useful tasks within years
“And I think in a few years robots will Start to do super useful stuff.”
Altman: Global demand for humanoid robots will outstrip supply chains
“The demand for humanoid robots in the world will be far more than we know how to think about with the current supply chain.”
Tan: US cannot produce precision screws or sheet metal without cost overruns
“You know, we can't even build precision screws and Large sheet metal without crazy cost overruns.”
Altman: AI and robotics enable bringing manufacturing back to the US
“And, you know, like, AI and robotics does give us a new possibility of a way to bring manufacturing back here and to bring sort of these complex industries here in a really important new way. And I would say that's at least worth trying.”
Altman: OpenAI is focused on making ChatGPT the best super-assistant
“We are going to try to make the best super assistant out of ChatGPT that we can. We're going to Add the things that we think we need to add to that but that is, like, one small part of the opportunity in front of us, and it makes us sad when people are like, I…”
Altman: OpenAI had no defensibility strategy for ChatGPT in its early days
“It took us a long time to figure out how to answer the defensibility question for ChatGPT. We had built this thing, and for a long time, the only defensibility was, like, we had the only product out in the market. And then we kind of, like, a brand that starte…”
Elon Musk emailed OpenAI saying it had zero percent success chance
“I can say this one because it got publicized in early, not early, a few years into OpenAI, where Elon sent us this really mean email. We'd been working through that for a while and said we had a zero percent chance of success. Like, not 0.10, that we were tota…”
Altman: AI models like o3 can act as junior employees executing multi-hour tasks
“But now you start to see things where you can like really give a task to Codex, for example or to deep research, and you have this thing go off and do a bunch of stuff and come back to you with like a proposal. It's like a very junior employee that can work on…”
Altman: AI enables the third major computing interface revolution, led by Jony Ive
“I think we've only had kind of two big revolutions in computer interfaces, really, in the last, like, 50 years. We had the, you know, keyboard and mouse and screen, and then we had touch and phones, and the opportunity to do a new one doesn't come along that o…”
Altman: Startups almost always win when industry clock cycles shift dramatically
“When the clock cycle of the industry changes this much, startups almost always win, and we've probably never seen it change this much.”
Altman predicts small teams will achieve step-function productivity gains from AI
“One of the things that will feel most different about
These next 10 years versus these last 10 years is how much a single person or a small group of people with a lot of agency can get done, and that is a bigger deal than it sounds like because coordination co…”
Altman: AI is somewhat mainstream, but most users only see basic chatbots
“I think AI is, like, somewhat mainstream right now. I, the negative, the way that it's not is most people still think of AI as ChatGPT, and a lot of people use ChatGPT, but they use it like a chatbot, and they have not yet wrapped their head around what's, wha…”
Altman: Never Looked at Colleges or Big-Tech Resumes for YC Applications
“When I was, like, reading YC applications, I would, like, never look at the resume items. You know, you worked at, like, Google, or went to this college, or I never cared. I would always go right to, like, what's the most impressive stuff you've done, and then…”
Altman predicts humanity will achieve unimaginable superintelligence within 20 years
“Look, there, there's a lot you know, in 10 or 20 years, unless something goes hugely wrong, we'll have, like, unimaginable super intelligence, and I'm very excited to see how that goes.”
Altman: Energy will eventually be the fundamental limiter on intelligence scaling
“I cannot recall ever thinking until like after starting OpenAI that they were going to be so obviously related. That, you know, the, that energy would be the eventually the fundamental limiter on how much intelligence we could have.”
Altman: AI progressed to PhD-level intelligence across most areas in five years
“In five years we have gone from this, like, thing that could barely write a sentence to a thing that is like, you know, PhD level intelligence in most areas.”
Altman predicts AI will maintain its rapid rate of progress for five more years
“Five more years I think we'll be able to maintain the same rate of progress”
Altman: AI's economic impact will be faster and steeper than the transistor's
“The analogy I like most for AI is the transistor, like the historical tech, technical analogy. You know, some people figured out like a new, really important scientific discovery. And society, the economy, whatever you want to call it, just got to work. Just d…”