Sam Altman Says the Next Trillionaire Is Sitting in the Room
- 5 hours ago
- 4 min read
Overview
• OpenAI CEO Sam Altman closed Y Combinator's Startup School 2026, held July 25–26 at Chase Center in San Francisco, drawing roughly 6,000 founders selected from 30,000-plus applicants — in a conversation with YC president Garry Tan titled "Never a Better Time to Do a Startup."
• Nvidia CEO Jensen Huang spoke at the same event, calling it "the greatest time ever to start a company" and saying he's "jealous of everyone in the room" — a compute-side CEO backing the same claim a model-side CEO was making from the same stage.
Altman's evidence isn't just enthusiasm: a coding agent produced, in roughly seven minutes, the same quantity of work it took him three months to write by hand building his 2005 startup, Loopt. The claim carries weight because OpenAI is the receipt — the company closed a $122B round in March 2026 at an $852B valuation and now runs roughly $2B a month in revenue, up from about $1B a quarter at the end of 2024.
The Room and the Moment
YC didn't scale Startup School down for 2026 — it moved the event to Chase Center, the Golden State Warriors' arena, and ran it over two days, July 25–26. Roughly 6,000 founders attended out of more than 30,000 applicants, about three times the size of the prior year's event. Altman closed the weekend in a fireside conversation with YC president Garry Tan under the session title "Never a Better Time to Do a Startup." Nvidia's Jensen Huang took the same stage and delivered a parallel verdict from the hardware side of the industry. Two CEOs sitting at opposite ends of the AI stack — one selling models, one selling the chips that run them — making the same claim inside the same 48 hours is itself a data point.
Three Months, Then Seven Minutes
Altman's case comes with a before/after, not just a slogan. In 2005, as a Stanford dropout building the location-sharing app Loopt, he spent roughly three months writing code by hand for the product's first version. Two decades later, he watched a coding agent produce an equivalent quantity of output in about seven minutes. That compression, three months down to seven minutes, is the basis for his headline claim about how much more a founder can now ship in the same calendar time.
"I kind of think there's some future trillionaire sitting in this room. I would bet that the startups created today will be much more valuable, much more impactful than startups of the past."
The Dystopia He's Actually Worried About
Altman didn't frame the acceleration as risk-free. Asked what could go wrong over the next decade, he pointed at overcorrection on safety rather than underinvestment in it — a surveillance-state outcome dressed up as protection.
"There's one dystopia that I'm particularly nervous about ten years from now. We overreact to AI safety and so we say, look everyone, you're gonna get a cure for cancer, you're gonna have material abundance but you will have no freedom, you will have no agency. It will be a perfect surveillance state."
Why "Everybody Was Calling Us Idiots" Was the Point
The contrarian-conviction advice lands harder given where OpenAI itself now sits. The company was founded in 2015 as a research lab widely dismissed as unrealistic; a decade later it closed the largest funding round in history, $122B in March 2026 at an $852B post-money valuation, with SoftBank co-leading and Nvidia, SoftBank, and Amazon each committing between $30B and $50B, months after an October 2025 employee tender sale worth $6.6B, in which more than 600 employees participated and roughly 75 hit the $30M-per-employee cap at a valuation near $500B at the time. Altman's advice to the room amounted to: bet on being wrong in public, the way OpenAI's early backers did.
"Find the things that you can develop reasonable conviction in that people decide the conventional wisdom is they're just wrong."
"Everybody was calling us idiots. That was an incredible gift."
The Unglamorous Advice
Underneath the trillion-dollar framing, Altman's tactical advice to the room skewed deliberately boring — play a long game, and ignore the internet's reward structure for cynicism.
"Be mildly helpful to many people over many years, with no immediate transactional expectation."
"Sarcasm is the opposite of what we like. It's very easy to take shots at people trying to do hard things and get 10,000 likes. Over the decade, none of those trolls did anything. Put all of your energy into building stuff."
The Closing Number
Altman closed on a pace claim, and OpenAI's own numbers back it up. The company's revenue run-rate went from roughly $1B a quarter at the end of 2024 to roughly $2B a month by early 2026, a growth curve OpenAI leaned on to justify the March funding round.
"The next six months will feel like the equivalent of the last two years of model progress."
"It's all going to work out. You can make a lot of mistakes. The tech industry is very forgiving. Have all the drive but be a little happier along the way."
Altman's capability argument is correct, and probably understated. Coding agents compressing three months of work into seven minutes is a real, measurable unlock, and OpenAI's own growth curve, a $2B-a-month revenue run rate and an $852B valuation less than a year after a $500B secondary, is the receipt.
Comments