Meta's $36 Billion VR Bet Failed, Now Repeating With AI Agents
Meta is building personal AI agents—autonomous systems that will handle tasks on your behalf, integrated into your phone, your messaging apps, your social graph.
Mark Zuckerberg announced it on the Q2 earnings call with the confidence of someone who has already decided the outcome. The company is committing serious resources to a future where you delegate decisions and errands to an AI that knows your habits, your contacts, your preferences. Your financial data.
This is not a feature—it's Meta's new platform bet. And it is an exact replay of a losing pattern the company has now executed twice.
In 2016, Facebook pivoted hard toward virtual reality, paying $2 billion for Oculus and then another $36 billion over the next seven years building out metaverse infrastructure. The vision was total and unambiguous—Zuckerberg appeared in earnings calls speaking about spatial computing the way he speaks now about agents. But Apple already owned VR through superiority in hardware integration. Microsoft already owned spatial productivity through enterprise relationships, so Meta entered a category where incumbents had already shaped user expectation and behavior and lost.
Most people haven't switched personal assistants once they settled into one.
”The structural similarity is not coincidental but a pattern baked into how Meta approaches platform transitions. The company does not enter new computing layers early—it enters late, at scale, with social graph integration as its differentiation strategy, betting that connectivity and mainstream reach can overcome the technical and trust advantage of entrenched competitors. This strategy failed with mobile, where Apple and Google had already won the operating system war before Meta could meaningfully participate. Now the company is entering personal agents late, behind Apple Siri, Google Assistant. OpenAI's full ecosystem, where the incumbent already owns the trust account and users already know what they are willing to delegate to these systems. Habit, not features, determines loyalty in autonomous delegation.
