VCs are responsible for managing billions of dollars of investments, but they can also be spectacularly wrong at times.
In September 2023, Keith Rabois — Founders Fund general partner and former executive at PayPal, LinkedIn, and Square — made a claim that has aged poorly with remarkable speed. “Not a single great SF startup has been founded since March 2020,” he had posted on X. In response, investor Sheel Mohnot had put out a public bet: $100,000 to the charity of the winner’s choice if any company founded in San Francisco between March 2020 and September 2023 reached a $5 billion valuation within ten years. When someone replied with a single word — “Anthropic” — Rabois was unmoved. “Highly skeptical,” he shot back. As it turns out, Anthropic is valued at $965 billion as of May 2026.
The exchange had context. Rohit Krishnan flagged Anthropic as the obvious candidate, noting it was already valued at around $4 billion at the time. Rabois brushed that off too, calling it “a joke non-financial valuation” — by which he meant, as another commenter clarified, that a VC funding markup isn’t a real valuation the way an acquisition or IPO is. Rabois doubled down: “Especially a strategic investor is definitely not a reflection of reality.”
It was a defensible position in a narrow technical sense. But as a read on Anthropic’s trajectory, it was about as wrong as a prediction can get.
Anthropic was founded in 2021 by Dario Amodei, Daniela Amodei, and a group of researchers who walked out of OpenAI over AI safety concerns. The company they built in San Francisco has since become one of the most consequential technology companies on the planet. Within two years of Rabois’ dismissal, Anthropic’s revenue had grown from $100 million in 2023 to $4.5 billion by mid-2025 — a 45x increase in eighteen months.
The valuation story is even more striking. In March 2025, Anthropic raised $3.5 billion at a $61.5 billion valuation. By September of that year, that figure had tripled to $183 billion. In February 2026, the company raised $30 billion at a $380 billion valuation. And just recently, it closed a Series H at $965 billion — making it, at least on paper, more valuable than OpenAI, the very company its founders had left.
The $5 billion threshold that Rabois was skeptical Anthropic would ever reach? The company blew past it before 2025 was halfway done.
What makes the original skepticism particularly worth revisiting is Rabois’ framing. His objection wasn’t merely about valuation methodology — it was a broader claim that the pandemic era had been creatively sterile in San Francisco, that trend-chasing founders with FOMO always fail, and that nothing great had emerged from the city since March 2020. Anthropic was, in his view, neither real enough nor good enough to count as a counterexample.
That framing has collapsed entirely. Anthropic’s revenue run-rate has since hit $47 billion, growing at roughly 10x annually. Its share of combined OpenAI-Anthropic business spending went from around 10% at the start of 2025 to over 65% by February 2026. Claude Code alone — the company’s agentic coding platform — has crossed $2.5 billion in annualized revenue, with weekly active users doubling in a matter of months.
There is a broader lesson here. Rabois’ skepticism about Anthropic reflected a common cognitive error among investors: confusing a funding round’s imperfections as a valuation signal with the underlying business being unreal. A VC markup is not the same as a public market price, but it is also not nothing. When a company at $4 billion is “one of the two leaders in the hottest and fastest growing area in tech,” as Rohit Krishnan put it at the time, dismissing it as a joke requires an unusually high degree of confidence in one’s own judgment.
Rabois is known for contrarian positions — on remote work, on hiring, on the geography of innovation. Some of those views are genuinely insightful. But the Anthropic call was a reminder that contrarianism, applied as a default posture rather than a considered analytical conclusion, is just as likely to produce a confident miss as a prescient insight.
The bet Sheel Mohnot offered was for $100,000. Based on where Anthropic sits today, that would have been the most lopsided wager in recent startup history.