AI industry insiders just delivered a surprising reality check. At the annual Cerebral Valley conference in San Francisco, more than 300 AI company founders, investors, and engineers participated in an anonymous survey that reveals growing skepticism about the industry's hottest companies. The results suggest a major shift in sentiment, with attendees betting against OpenAI while predicting massive wins for Nvidia.
The annual Cerebral Valley conference just delivered the kind of insider intelligence that moves markets. Built around candid conversations between AI's power players, this year's event featured something unprecedented: an anonymous survey of more than 300 attendees that cuts through the industry's typical cheerleading to reveal what insiders really think.
The survey participants weren't random observers. They were primarily AI company founders, followed by investors, product leaders, engineers, and media covering the space. When The Verge's Alex Heath shared the results onstage, the numbers painted a picture of an industry recalibrating its expectations.
OpenAI, despite Sam Altman's projection of $20 billion in annualized revenue this year, faces tempered enthusiasm from the very people building the AI ecosystem. Insiders predict the company will reach $30 billion by 2026 - solid growth, but not the explosive trajectory many expected. More telling: when asked which private companies they'd invest in today, Anthropic topped the list, ahead of OpenAI.
The skepticism runs deeper. Perplexity, the AI search startup, dominated the 'companies you'd short' category among billion-dollar valuations. OpenAI came in second on that same list, a striking vote of no confidence from industry peers. As one anonymous attendee put it during breakout sessions, these companies have "infinite money" but finite talent - and the race is getting brutal.
Nvidia emerges as the survey's biggest winner, with insiders predicting a $6 trillion valuation by 2026. That's nearly double its current market cap and reflects the chip giant's stranglehold on AI infrastructure. While AI companies struggle with business models, Nvidia keeps selling the picks and shovels.
The most revealing shift involves AGI timing. Survey respondents don't expect artificial general intelligence to be declared until 2030, pushing back the timeline that dominated earlier conferences. This reflects what Heath observed throughout the day: "No one cares about AGI anymore." The focus has shifted from existential promises to immediate business applications.
Meta faces particular challenges according to the survey. The company wasn't mentioned among models likely to lead LM Arena rankings by 2026, despite its significant Llama investments. Instead, Chinese model Qwen from Alibaba cracked the top five, signaling how quickly the competitive landscape is shifting. Many companies are fine-tuning open-source Chinese models rather than Meta's offerings.
The survey reveals other fascinating dynamics. Andreessen Horowitz topped the list of VC firms with the most enviable AI portfolios, followed by Khosla Ventures and Sequoia. When it comes to model leadership next year, insiders still give OpenAI the edge, followed by Anthropic and Google's Gemini.
Beyond the numbers, the conference atmosphere reflected an industry maturing rapidly. Gone are the days of founders predicting humanity's demise by GPT-10. Instead, as Replit CEO Amjad Masad noted onstage, "If you are competing on price, then maybe you don't have a business." The message is clear: build something people will pay for.
San Francisco Mayor Daniel Lurie captured the broader sentiment: "People are starting to complain about traffic. Thank goodness. I want those complaints." The AI boom is creating real economic activity, not just venture hype.
The reverse acquihire trend also drew significant discussion. Meta's deals with ScaleAI and Google's acquisitions of Character and Windsurf reflect what attendees see as a talent war with unlimited budgets. One AI founder recalled a Big Tech corporate development team asking him to name his own valuation.
Perhaps most importantly, the survey reveals an industry grappling with bubble fears while staying focused on fundamentals. As Anthropic CPO Mike Krieger emphasized, "Time spent is not on any of the dashboards that I look at." The focus is shifting from engagement metrics to sustainable business models.
The Cerebral Valley survey offers a rare glimpse into what AI's builders really think when the cameras aren't rolling. Their predictions suggest an industry shifting from hype to fundamentals, where even OpenAI faces scrutiny and practical applications matter more than AGI timelines. For investors and companies betting on AI's future, these insider perspectives provide crucial signals about where the smart money is really going - and which companies might be riding for a fall.
