Ilya Sutskever: The OpenAI Co-Founder Who Tried to Fire Sam Altman, Then Built a $32B Company
He co-created the deep learning revolution, led the failed coup at OpenAI, apologized publicly, resigned quietly, and raised $3 billion for a company with zero revenue and 20 employees.
The Key Lesson
"Reputation is the ultimate currency. Ilya raised $3 billion for a company with no product, no revenue, and 20 employees. When you've helped create an entire field, investors bet on your vision, not your spreadsheets."
The Architect of Deep Learning
Before the drama, before the billions, Ilya Sutskever was simply one of the most important researchers in artificial intelligence history. Born in Russia, raised in Israel, educated in Canada under Geoffrey Hinton—the "godfather of AI"—Ilya helped build the foundations of modern deep learning.
In 2012, Sutskever co-authored the AlexNet paper with Hinton and Alex Krizhevsky, proving that deep neural networks could dramatically outperform traditional computer vision approaches. That paper sparked the entire deep learning revolution.
When OpenAI launched in 2015, Ilya was there as co-founder and Chief Scientist. For nearly a decade, he led the research that made OpenAI the world's most valuable AI company.
The Week That Shook Silicon Valley
In November 2023, Ilya Sutskever made a move that shocked the tech world. As a board member of OpenAI, he authored a 52-page memo accusing CEO Sam Altman of lying, manipulating executives, and fostering internal division. The board fired Altman.
The memo relied heavily on information from Mira Murati, OpenAI's CTO. It was a calculated attempt to wrest control of the company from its charismatic CEO.
It lasted about 72 hours.
Sutskever expected a muted reaction from staff. Instead, hundreds of employees threatened to quit unless Altman was reinstated. Microsoft, OpenAI's biggest investor, made clear where they stood. Within days, Altman was back, and Sutskever was apologizing publicly.
The Quiet Exit
In May 2024, Ilya announced his departure from OpenAI. He didn't make a scene—just a statement about pursuing "something personally meaningful."
Jan Leike, who co-led OpenAI's superalignment team with Sutskever, announced his departure hours later. He was more explicit, citing "an erosion of safety and trust in OpenAI's leadership."
The failed coup had exposed a fundamental rift: Ilya believed OpenAI was moving too fast, prioritizing products over safety. He lost the battle, but he hadn't given up the war.
Safe Superintelligence Inc.
In June 2024, just weeks after leaving OpenAI, Ilya announced Safe Superintelligence Inc. (SSI). The company's mission? Build superintelligent AI safely. Its first product? Superintelligence itself.
"Our first product will be the safe superintelligence, and it will not do anything else up until then. Ultimately, I had a big new vision. And it felt more suitable for a new company."
— Ilya Sutskever
He co-founded SSI with Daniel Gross (former Y Combinator partner) and Daniel Levy, establishing offices in Palo Alto and Tel Aviv.
$1B (Sept 2024)
First raise at $5B valuation
$2B (Mar 2025)
Second raise at $32B valuation
~20 Employees
Pure research focus
$3 Billion, Zero Revenue
In September 2024, SSI raised $1 billion from Andreessen Horowitz, Sequoia Capital, DST Global, and SV Angel—at a $5 billion valuation. The company had no product, no revenue, and approximately 20 employees.
By March 2025, they'd raised another $2 billion, reaching a $32 billion valuation. Still no product. Still no revenue. Still around 20 employees.
What they did have was Ilya Sutskever—and that was enough.
Meta Wants In, Ilya Says No
In the first half of 2025, Meta attempted to acquire SSI. The offer was rebuffed. Ilya had built SSI to do one thing—create safe superintelligence—and he wasn't about to let it become a division of a social media company.
Google Cloud announced a partnership to provide TPUs for SSI's research. The hyperscalers wanted to be in Ilya's orbit, even if they couldn't own him.
The Philosophy
SSI represents Ilya's thesis about what went wrong at OpenAI. Instead of releasing products, chasing revenue, and racing to ship features, SSI is focused purely on research. They're not building ChatGPT competitors—they're trying to solve alignment before it's too late.
It's a bet that the company that figures out how to make AI safe will ultimately win the entire market. And Ilya is willing to forgo billions in potential revenue to prove it.
What Founders Can Learn
Reputation compounds over decades
Ilya spent 15+ years building credibility in AI research. That reputation raised $3B with no product. Your track record is your greatest asset.
Political moves can backfire spectacularly
The OpenAI coup failed because Ilya misread the room. Even brilliant researchers can be blind to organizational dynamics. Know your stakeholders.
Failure doesn't have to be final
Ilya publicly failed at the biggest company in AI. Months later, he raised billions for a new venture. One failure doesn't define you.
Sometimes you need to start over
"It felt more suitable for a new company." When your vision doesn't fit the existing structure, sometimes the only path forward is building something new.
Conviction attracts capital
SSI's pitch is essentially "trust us to build the most important technology in history." That level of conviction—backed by credibility—attracts believers.
The Long Game
Ilya Sutskever isn't racing to ship products. He's not optimizing for ARR or user growth. He's playing a longer game—one where the winner is whoever figures out how to make superintelligent AI safe and beneficial.
The failed coup at OpenAI could have ended his career. Instead, it became the catalyst for something bigger. Sometimes you have to lose one battle to start a war you can actually win.
With $3 billion in the bank, 20 researchers, and no product deadlines, Ilya has the runway to think in decades. Whether SSI succeeds will determine a lot more than the future of one company.
It might determine the future of humanity.
Write Your Own Story
See what VCs find when they research your startup. AI-powered due diligence — free in 60 seconds.
Get Free AI Analysis