The Metaverse

Strip away the keynote polish and the question was simple: what could you actually do in there? The honest answer, for most people, was not much they wanted to. The flagship experience, Horizon Worlds, put you in a cartoon world as a legless avatar floating from the waist up. Legs were, for a long time, a running joke — the technology to track them convincingly wasn't ready, so avatars simply didn't have any. It's a small detail that captured a large gap between the promise and the thing.

The hardware carried real friction. A good headset was heavy, warm, and isolating; wearing one for an hour left many users with eye strain or mild nausea. It cut you off from the room and the people in it, which is a strange foundation for a technology sold on togetherness. And the social worlds were often empty. Internal targets for Horizon Worlds reportedly aimed for millions of monthly users; leaked figures suggested it was struggling to hold a few hundred thousand, many of whom didn't come back after the first month.

The unwinding didn't come as a single announcement. It came as a slow change of subject. Through 2022, Reality Labs kept posting enormous losses while the rest of Meta's business softened, and the stock fell hard — Meta shed roughly two-thirds of its value that year before recovering. Investors who had tolerated the metaphysics started asking blunt questions about the billions going into a division most of them couldn't picture using.

Zuckerberg's public framing shifted with the mood. In early 2023 he declared it Meta's "year of efficiency," a phrase that meant layoffs — the company cut around 21,000 jobs across two rounds. He didn't renounce the metaverse, exactly. He just stopped leading with it. Then came the arrival of generative AI at the end of 2022, and with it a new story that was cheaper to tell, easier to demo, and immediately useful. On earnings calls through 2023, mentions of "AI" overtook "metaverse" so completely that outside observers counted the words as a kind of weathervane.

What's striking about the metaverse isn't that it faded. It's that it had faded before, under other names, and will almost certainly return under the next one. The dream of a shared, embodied, three-dimensional internet is older than Meta's rename by decades — Second Life in 2003, virtual-reality hype in the early 1990s, the cyberspace of the same Stephenson novel that gave us the word. Each wave arrived declaring the future had begun. Each receded when the hardware and the human appetite didn't cooperate. Meta's version was the biggest-funded and most public, which is why its retreat felt like an event rather than a footnote.

This points to a pattern in how technology gets sold: the name arrives before the thing. Naming a future does real work — it organizes capital, aligns competitors, gives investors a story to price. But it also front-loads expectation onto a product that doesn't exist yet, which sets up the crash. The metaverse was announced as a destination when it was, at best, a direction. And a direction sold as a destination will always look like a failure when people arrive and find a construction site.

Three years after the rename, Meta is still called Meta, still ships headsets, and still pours billions into Reality Labs — but almost no one describes it as a metaverse company anymore, including Meta. The word did its job and then became a liability, and the company moved its energy to a story with faster returns. Horizon Worlds still exists. The avatars have legs now. Hardly anyone noticed, because the conversation had already gone somewhere else.