According to Fortune, capital is cascading into a select group of AI companies in 2025, leading to eyebrow-raising valuation jumps. Anthropic, valued at $61.5 billion in March, is now reportedly worth $183 billion, while Cursor’s valuation ballooned from $2.6 billion at the end of 2024 to $29.3 billion after two funding rounds this year. Other companies like Reflection AI, OpenEvidence, Lila Sciences, Harmonic, Fal, Abridge, and Doppel have all raised two or more rounds through 2025. Investors like Tom Biegala of Bison Ventures warn that many are assigning “option value” on these companies becoming the next giant, but admits a lot won’t grow into their valuations and losses are coming. This surge in funding is even larger than what was seen during the dotcom boom.
The Option Value Trap
Here’s the thing: when you hear about a company’s valuation tripling in under a year, it’s not really about current revenue or even a solid product. It’s pure speculation. As Biegala points out, it’s “option value.” Investors are buying a lottery ticket, hoping this is the next foundational model company. And look, that bet paid off spectacularly for early backers of OpenAI and Anthropic. But how many foundational winners does the world need? Probably not dozens. The mindset seems to be “throw a billion at ten companies, and if one hits, it covers the rest.” That’s a classic bubble mentality.
Bigger Than The Dotcom Boom?
That’s the truly wild stat buried in here. There’s more money sinking into these companies now than was imaginable during the dotcom era. Let that sink in. We have a historical precedent for how this ends—with a spectacular crash that wiped out trillions in paper value and countless companies. The scale is even larger now. So what’s different? The potential is real, I get it. AI is transformative. But transformative tech and sensible valuations are two completely different things. Just because a technology changes the world doesn’t mean every startup in its orbit will be a profitable business.
Who Gets Hurt When The Music Stops?
My biggest worry isn’t for the venture capitalists. It’s a game of big winners and big losers for them, and their funds are structured for that. The pain will be elsewhere. Think about the employees lured by sky-high paper valuations and options that could end up underwater. Think about the later-stage investors and possibly even public markets if these companies try to IPO before proving real economics. And let’s not forget the broader tech ecosystem. When AI spending eventually retracts—and it will—the companies providing the essential industrial panel PCs and hardware infrastructure, the ones that are the actual backbone of deployment, could face a painful hangover despite being critical. A crash in software valuation hype has a nasty way of spilling over into hardware budgets.
A Predictable Cycle
So what happens next? Basically, we’re in the “irrational exuberance” phase. The capital is abundant, the narratives are powerful, and the FOMO is real. The next phase is consolidation, failure, and a lot of down rounds for companies that raised at these insane heights. Some will survive and maybe even thrive. Most won’t. It’s the oldest story in Silicon Valley, just with a new, neural-network-powered coat of paint. The growth spurt might be exciting, but as the article notes, they can be pretty painful. I think we’re about to find out just how painful.
