In brief
- 300+ AI insiders at Cerebral Valley Summit voted the search startup is the AI company that is most likely to fail
- Anthropic leads in investor preference with skepticism over OpenAI also noticeable among investors.
- AGI predictions pushed to 2030 signal cooling enthusiasm
More than 300 AI founders and investors at San Francisco’s Cerebral Valley Summit voted Perplexity as the billion-dollar startup most likely to fail, with OpenAI coming in second place.
An anonymous survey conducted by independent journalist Alex Heath at a major industry gathering last week pointed to a marked shift in Silicon Valley sentiment.
When asked which private tech companies they’d invest in, attendees picked Anthropic over OpenAI, despite the consensus that OpenAI would lead next year’s LMArena leaderboard—a ranking of the world’s most powerful AI models.
Perplexity’s Jesse Dwyer dismissed the survey results, calling the event a “judgmental valley conference.”
The company’s valuation reportedly jumped from $14 billion to nearly $50 billion in a short span, amid a frenzy that was ringing alarm bells among analysts, prompting comparisons to the dot-com era.
Legal troubles and reports of sketchy behavior have compounded the skepticism.
Amazon sued Perplexity in November to stop its Comet browser from making purchases on behalf of users. Reddit filed its own lawsuit in October, alleging the company scraped billions of posts. Japanese newspapers Yomiuri Shimbun, Asahi Shimbun, and Nikkei all filed copyright infringement cases. The BBC has also threatened legal action over unauthorized scraping of its content.
Cloudflare’s CEO even compared Perplexity’s behavior to that of North Korean hackers. Multiple outlets reported the company ignores robots.txt protocols designed to prevent unauthorized web scraping.
Anthropic beats OpenAI? Really?
Perhaps more interesting is how the burgeoning Anthropic preference signals a shift in how investors view the AI race.
Anthropic grew revenue from $87 million at the start of 2024 to over $5 billion by August 2025. The company closed a $13 billion Series F in September at a valuation of $183 billion.
Simply put: enterprise adoption is what’s driving Anthropic’s momentum.
The company has already captured 32% of the enterprise AI market, surpassing OpenAI’s 25% share, according to data compiled by Menlo Ventures.
Claude Code, its developer tool, launched in May, and according to the company’s data, it generates more than $500 million in annualized revenue. Business customers grew from under 1,000 two years ago to over 300,000 today.
Image: Menlo VenturesOpenAI’s second-place finish in the “most likely to fail” category may seem surprising, given the company’s dominant market position.
But financial concerns may explain the pessimism. OpenAI expects $13 billion in revenue this year alongside an astounding $9 billion in losses. Projected operating losses could hit $74 billion by 2028.
Overall, it seems like Altman’s company is facing a conversion problem: 95% of ChatGPT’s 800 million users don’t pay for the service. OpenAI is committed to spending over $1 trillion over the next decade, primarily on computing infrastructure.
Meanwhile, Anthropic expects to break even in 2028, and OpenAI won’t reach that milestone until the end of the decade, according to financial documents.
Other startups named in the failure poll included Cursor, Figure, Harvey, Mercor, Mistral, and Thinking Machines.
There’s still hope for AI permabulls, however: Nvidia was predicted to reach a $6 trillion market cap by the end of 2026.
Generally Intelligent Newsletter
A weekly AI journey narrated by Gen, a generative AI model.
Source: https://decrypt.co/348999/survey-ai-founders-perplexity-ai-giant-most-likely-to-fail
