In what felt like a candid moment of Silicon Valley self-critique, more than 300 founders, investors, and AI execs were asked at the Cerebral Valley AI Conference: “Which billion-dollar startup would you bet against?” The answer that stood out? Perplexity — the AI search engine startup that’s raised at increasingly stratospheric valuations and now finds itself at the top of the list of companies most likely to crash. Surprisingly, even OpenAI, long seen as the safe bet in the AI boom, came second in the informal poll. The assessment is unscientific, anonymous and unconventional — but it carries weight precisely because the players don’t often publicly bet against each other. In that sense, this vote isn’t about burying one startup; it’s a symptom of the anxiety rippling through the AI industry.
Here’s what’s really going on behind the headline: Perplexity is pulling high valuations (as high as $14 billion to a rumored $50 billion), but many at the conference questioned its business footing. The skepticism: technical promise is there, but product maturity, monetization and sustainable moat? Not so clear. That combination makes it a lightning rod for doubts about how many generational AI companies can truly scale. Even OpenAI’s massive scale and marquee role haven’t insulated it from scrutiny: its enormous infrastructure commitments and ballooning cost structure have raised concerns about whether the business model matches the hype.
What this moment underscores is threefold. First, AI isn’t immune to startup risk. The idea that every company raising big valuation, building large models, or flashing enterprise logos will succeed is being challenged. Second, the crowd’s willingness to single out a major player suggests that investors and founders believe some of the high-valuation bets may not pay off — even in a red-hot market. That doesn’t mean collapse is inevitable, but that belief in unlimited upside is softening. Finally, for entrepreneurs and investors, this vote serves as a reminder: your narrative matters, but so does the business underneath. Hype will get you attention; execution will — or won’t — get you survival.
For you watching from the sidelines: Maybe you’re evaluating an AI vendor, a startup job, or investment in the space. One takeaway is to ask beyond “How big can this be?” and probe “What happens if the world doesn’t adopt it as fast as the valuation assumes?” Look at how companies defend themselves, whether they have real customers paying, how predictable their economics are, and how they’ll handle a change in climate. Because in an era of AI goldrush, the questions about who won’t succeed matter just as much as who will.
In short, the conference ballot may have felt like a fun gimmick, but its implication runs deeper: even in a field full of optimism, doubt is returning — and the startups at the front of the line are increasingly aware they could be the next cautionary tale.
