When no one shows up in a virtual room, there’s a certain silence. That is, in many respects, the epitome of Meta’s metaverse era: sophisticated press releases, intricate digital architecture, and a nearly empty interior. When word started to spread that Meta was pulling out of its flagship virtual reality platform, Horizon Worlds, the company had already invested more than $80 billion in the project. Furthermore, the general public was still unaware of what a metaverse was.
It’s important to consider how forcefully Zuckerberg marketed this idea. He did more than simply make a product announcement in late 2021. One of the most well-known businesses on the planet was renamed by him. Facebook evolved into Meta. It was a proclamation, not a change in direction, that the next phase of human interaction would take place in virtual reality while wearing goggles and gazing at blocky, legless digital avatars. Workers at the time reported feeling under pressure to use internal virtual reality tools for meetings. According to reports, many resisted. The avatars drifted. In any case, the meetings took place via Zoom.
The cost of this experiment was borne by Meta’s Reality Labs division. Since 2021, the unit—which houses both the hardware and software for the metaverse, such as Quest headsets and Ray-Ban AI glasses—has reported cumulative losses exceeding $70 billion, with over $19 billion in losses in 2025 alone. These are not rounding mistakes. They are the price of creating something that the market never really demanded, at least not at the speed or scale that Zuckerberg had in mind.
The story that Meta just burned $80 billion and left is, to be fair, a little deceptive. Reality Labs is not going to close. There seems to be some real consumer interest in the hardware side, especially with regard to the Ray-Ban smart glasses. Filmmaker James Cameron is reportedly one of the enthusiasts for a new pair of glasses that are being developed for home streaming. The effort is ongoing, though it’s still unclear if any of this will be significant at scale. The funds did not disappear into thin air; rather, they established a solid basis for Meta’s upcoming hardware endeavors.
Zuckerberg’s public enthusiasm for the metaverse concept itself undoubtedly waned. His current vocabulary hardly contains the word. Today’s announcements focus on AI infrastructure, large language models, and what Meta is framing as the pursuit of “superintelligence.” In contrast, 2021 press conferences were rife with metaverse terminology. It’s difficult to miss the whiplash. Within about three years, he went from declaring billion-dollar investments in data centers and AI talent to promising that people would teleport as holograms to attend family dinners.
The metaverse wager doesn’t seem to have been wholly irrational at the time. There was real momentum for augmented and virtual reality.
It was referred to as a trillion-dollar opportunity by investment firms. Barbados established a physical embassy within Decentraland, a metaverse world that now reads more like a time capsule of 2021 hype than like progressive diplomacy. Not only was execution flawed, but the fundamental value proposition never resonated with the general public. There was nothing that a phone call or video chat couldn’t accomplish more comfortably than putting on a headset to attend a meeting or mingle in a pixelated room.
The idea that the metaverse chapter is just over has been refuted by Meta’s leadership. They contend that social and immersive experiences can develop through phones, glasses, and other unidentified surfaces in the metaverse, negating the need for headsets. Although it conveniently broadens the definition enough to make failure more difficult to measure, it is a reasonable reframe.
Observing this over a four-year period, the speed of the pivot is more noteworthy than the magnitude of the loss. The mobile internet was meant to be replaced by the metaverse. Almost instantly, a new, all-consuming priority took its place, turning it into a footnote. The question at the heart of Meta’s story is whether AI can deliver where the metaverse failed, and for the next $80 billion, the answer probably matters more than anyone is letting on.
