The $80 Billion Ghost World

Mark Zuckerberg thought he built the successor to the internet. He delivered a realm of legless avatars, empty virtual plazas, and the most expensive vanity project in Silicon Valley history.

As the stats sit right now, the metaverse has plunged: $80 billion lost in Reality Labs since 2020, fewer than 200K monthly users at its peak versus the proposed 1 billion users and 5 years from the grand rebrand to the total shutdown.

Silicon Valley has an issue, and the issue is hubris – the belief that if you simply declare a future into existence with enough conviction and enough capital, reality will eventually comply. Mark Zuckerberg believed this with every fibre of his being. He was wrong in ways that will be studied for decades.

In October 2021, Zuckerberg stood before the world and announced that Facebook would henceforth be known as Meta. The metaverse, he declared, was "the next frontier". Not a product. Not a feature. A frontier. The very successor to the mobile internet. Within a decade, a billion people would live, work, and play inside it.

"An empty world is a sad world."

Internal Meta document on Horizon Worlds

Horizon Worlds, the flagship platform meant to be the beating heart of this digital civilisation, had been removed from Quest headsets. The app was scrubbed from the store. The metaverse was over. The receipt: nearly $80 billion in losses from Reality Labs alone, over $6 billion torched in just the final quarter of 2025.

The Cyberpunk Dream That Stayed Cartoonish

The vision was borrowed from science fiction — from William Gibson's Neuromancer, from Neal Stephenson's Snow Crash, and from the shimmering neon corridors of Ready Player One. A persistent, shared virtual space where presence would feel real, where economies would flourish, where human connection would transcend physical geography. It was an intoxicating pitch.

What materialised was… an empty room. Horizon Worlds launched to the public and immediately became an object of ridicule. Zuckerberg's own blocky avatar floating before crude 3D landmarks on a featureless green plane went viral for all the wrong reasons. The avatars had no legs. The worlds were empty. Less than 1% of users ever built anything inside the platform; most of the worlds that were built were never visited by more than 50 people. Internal documents, later leaked, carried the damning verdict in plain language: an empty world is a sad world.

Most users never came back after their first month. The internal target of 500,000 monthly active users was revised down to 280,000, which was itself never met. At its peak, Horizon Worlds counted fewer than 200,000 monthly active users.

• 

• 

• 

A Miscalculation of Epic Proportions

The fundamental error was not technological; it was anthropological. Zuckerberg assumed that if you built an immersive virtual world and called it the future, humans would adapt their behaviour to inhabit it. They did not. To enter the metaverse, a user needed to purchase a VR headset costing hundreds of dollars, strap it to their face, and actively choose to spend time inside. The barrier was financial and existential. Nobody wanted to live in a cartoon.

While Meta was burning billions constructing digital cathedrals no one visited, a text box called ChatGPT launched in November 2022 and broke the internet. The future was not virtual worlds and realities; the future was artificial intelligence. Threads attracted more users in its first hour than Horizon Worlds managed in two and a half years. (Conor Murray: Threads, Meta’s X Competitor, Hits 500 Million Monthly Users). OpenAI delivered immediate, tangible value. The metaverse delivered a ghost world with bad graphics.

A Timeline of Hubris

2021

Facebook becomes Meta; Zuckerberg bets the company on the metaverse, promising a billion users within a decade and "hundreds of billions in digital commerce".

2022

The cracks appear; Horizon Worlds launches publicly. The infamous legless-avatar screenshot goes viral. Internal documents reveal fewer than 200,000 monthly users and a user base that won't return after the first visit.

2023

The pivot begins; Zuckerberg starts reframing Meta around AI. Reality Labs losses surpass $13 billion for the year. The word "metaverse" begins its slow disappearance from official communications.

2024

Silence: The word "metaverse" vanishes entirely from the third-quarter earnings calls. Global VR headset shipments fall 12% — the third consecutive year of decline. The Ray-Ban AI glasses become Meta's only real hardware win.

2026

The end: Horizon Worlds is removed from Quest headsets in March. A brief reversal days later changes nothing. $80 billion and five years after the grand rebrand, the metaverse is officially dead.

What It All Means

The metaverse failure is not simply a story about the limits of techno-determinism — the Silicon Valley theology that insists technology creates behaviour, rather than serving it. Zuckerberg did not build the metaverse because users wanted it. He built it because he feared being on the wrong side of the next platform shift, the way Facebook had been caught flat-footed by cellphones and apps such as Instagram, Twitter and TikTok.

The irony is that in learning the wrong lesson, he may simply be repeating the mistake at a larger scale. Meta's $80 billion is now chasing AI because Zuckerberg is terrified of missing another wave. Only the platform has changed.

What the metaverse leaves behind is less a legacy than a cautionary tale: that you cannot rebrand reality. You cannot spend your way into the future. And you cannot convince a billion people to strap a headset to their face and inhabit a world of legless cartoons just because you've renamed your company after it. The future, it turns out, has its own opinions.