Chamath Palihapitiya, the former Facebook vice president who helped drive the platform from a college network to a global utility, has a habit of drawing straight lines between past warnings and present failures. He did it again this week on The Axios Show, where he accused Mark Zuckerberg and Meta of blowing the generative AI race — and compared the fumble to one he saw coming a decade ago. Palihapitiya told Axios he once tried to persuade Facebook to build its own smartphone platform.
He believed the company could have carved out a strategic position in mobile hardware, controlling the operating system that billions of people would use to access its services. Facebook passed.
Apple and Google locked in the mobile duopoly instead. Facebook became a tenant on their platforms, paying for access to its own users. The AI story, Palihapitiya argued, is the same pattern with different hardware.
This time the prize is the computing infrastructure that powers large language models and generative applications. And this time the winner is Nvidia, whose graphics processing units have become the de facto engine for training and running AI systems.
Jensen Huang, Nvidia’s chief executive, has seen his company’s market value soar past $3 trillion as demand for its chips outstrips supply. “They completely fumbled,” Palihapitiya said of Meta’s leadership. He stressed that he no longer has insight into the company’s internal decision-making.
“I don’t know the organisation well enough, nor do I understand the political dynamics to understand why they failed as miserably as they have. But they’ve profoundly failed.” The criticism carries weight because Palihapitiya was not a peripheral figure. He joined Facebook in 2007, when the company was still a startup with fewer than 100 million users.
He led the growth team that pushed that number past 700 million, refining the viral mechanics that made Facebook sticky and addictive. He left in 2011, after the company’s IPO filing, and later became a venture capitalist at Social Capital.
He has been publicly critical of social media’s harms — he once called the tools he helped build “ripping apart the social fabric” — but he has rarely aimed his fire directly at Zuckerberg’s strategic judgment. Meta has not been idle in AI. The company poured resources into fundamental research, releasing open-source models like LLaMA that rival proprietary systems.
It has deployed AI across its advertising business, its recommendation algorithms, and its content moderation pipeline. Zuckerberg has said Meta is investing heavily in compute infrastructure and plans to build its own custom silicon for AI workloads.
None of that, in Palihapitiya’s view, changes the central fact: Nvidia got there first. The company’s CUDA software platform, built over a decade, made its chips the default choice for developers. Huang bet early on accelerated computing and on the idea that the same processors used for graphics could be repurposed for scientific and AI workloads.
That bet paid off when the generative AI boom hit, and Nvidia was the only supplier with a mature, production-ready stack. Palihapitiya did not offer a theory for why Meta missed that window. He said he cannot explain the internal dynamics that led to the outcome.
But his comparison to the smartphone platform episode suggests he sees a recurring blind spot: a tendency to treat hardware as an afterthought, as something to buy rather than something to own. Facebook once had the users, the data, and the engineering talent to build its own mobile operating system.
It chose not to. Meta had early access to the AI research community, a massive trove of training data, and the financial resources to build its own chip supply chain. It let Nvidia set the standard instead.
Palihapitiya’s critique is not a technical one. It is a strategic one.
And it echoes a question that has dogged Meta for years: whether the company’s leadership can see around corners, or whether it is always reacting to moves others have already made.
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “Former Facebook VP Says Meta Missed AI Opportunity He Predicted Years Ago”, “description”: “Chamath Palihapitiya, the former Facebook vice president who helped drive the platform from a college network to a global utility, has a habit of drawing…”, “datePublished”: “2026-06-26T08:09:00+00:00T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Erika Johnson”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.infopulsetoday.com/”}}
