Meta Killed the Metaverse and Bet Everything on AI. Sound Familiar?

On October 28, 2021, Mark Zuckerberg stood next to a legless virtual avatar and told the world Facebook was now Meta.

He said the metaverse would reach one billion people within a decade, host hundreds of billions of dollars in digital commerce, and create jobs for millions of creators and developers.

Metaverse was the hottest idea on the planet on 2021.

Microsoft announced metaverse Teams. NVIDIA launched Omniverse. Nike opened a virtual Roblox store. Nobody wanted to be left behind.

Four and a half years later?

That billion-user target is still a fantasy.

On March 17, Meta quietly posted on a community forum: the VR version of Horizon Worlds shuts down June 15. The app gets pulled from Quest headsets. The virtual world goes dark. A mobile version survives.

The numbers behind it are brutal. Reality Labs, the division that built all of this, has racked up nearly $90 billion in operating losses over seven years. Last quarter alone, $6 billion lost on less than $1 billion in revenue. The income didn't even cover a sixth of the losses.

In January, the division laid off over 1,000 people, shut down multiple VR studios, and killed nearly every virtual-world project still in development. By mid-March, Reuters reported Meta was planning to cut roughly 20% of its entire workforce, around 15,000 people. If it goes through, it would be the company's biggest layoff since 2022.

Where's the money going instead? AI.

Meta's capital expenditure budget for the year sits between $115 billion and $135 billion, nearly all of it earmarked for AI infrastructure.

Shut down the virtual world. Cut a fifth of the staff. Pour every dollar saved into AI.

The day the news dropped, Meta's stock rose 3%.

$90 Billion vs. $2 Billion

But the real story is the contrast.

One day before Horizon Worlds' death was announced, Manus, the AI agent company Meta acquired for $2 billion, launched its desktop version. It includes a feature called "My Computer" that lets AI step off the cloud and onto your local machine: reading files, opening apps, running terminal commands.

The Horizon Worlds experience: spend a few hundred bucks on a Quest headset, strap it on, adjust the fit, draw a safety boundary, enter a cartoon lobby where everyone floats around because nobody has legs. Play mini-games. Chat with strangers' avatars.

After 30 minutes, the headset digs into your face. After an hour, some people start feeling sick. Meta never once disclosed how many people actually used it.

The Manus experience: download an app, open it, type a sentence. "Organize the thousands of files in my Downloads folder by type." It scans your hard drive, creates subfolders, sorts everything. You never touch the keyboard.

In one demo, someone asked it to build a macOS app from scratch using their local dev environment. It took 20 minutes.

Manus had only been live for eight months before the acquisition. It already had over a million paying users and was on track for more than $100 million in annualized revenue.

One spent $90 billion inviting you into a virtual world, and nobody showed up, while the other spent $2 billion walking onto your actual desktop, and it's already making real money.

Picking the Right Direction Doesn't Mean the Road Is Smooth

If you only read headlines, Meta looks like a company that can't stop tripping.

Its flagship AI model, codenamed Avocado, was supposed to ship in March. Internal tests showed it lagging behind Google, OpenAI, and Anthropic on reasoning and coding, so the launch got pushed to May.

The previous generation, Llama 4, landed with a thud last year and barely registered in developer communities. Reports suggest Meta even discussed temporarily licensing Google's Gemini to fill gaps in its own products. A company spending $135 billion on AI infrastructure, borrowing someone else's model.

Chief AI Scientist Yann LeCun left to start his own company. Alexandr Wang, the new AI lead poached from Scale AI in a $1.43 billion deal, hasn't delivered results yet.

A 20% layoff, a metaverse shutdown, and a delayed model launch, all in one week. Looks like a company that doesn't know what it wants to be.

Except Everyone Else Is Doing the Exact Same Thing

In February, Block CEO Jack Dorsey announced 4,000 layoffs, nearly half the company. His memo didn't sugarcoat it: AI tools have changed how companies are built, and smaller teams can do more. The stock jumped 25% that night.

Shopify's CEO sent a company-wide memo with a new rule: before you request a new hire, prove AI can't do the job first.

Amazon cut 16,000 roles in January. Atlassian laid off 1,600 and pledged to redirect everything toward AI enterprise software. In the first 74 days of 2026, 166 tech companies collectively cut nearly 56,000 jobs.

Sound familiar? It should.

2021 played out the same way. After Zuckerberg rebranded as Meta, Microsoft pledged metaverse Teams. NVIDIA rolled out Omniverse. Nike opened a virtual store. Disney created a metaverse division. Shanghai and Seoul published metaverse strategy roadmaps. Everyone chased the same trend. Everyone was terrified of being left out.

Five years later, the trend changed. The behavior didn't.

Last time, the consensus was "the metaverse is the next computing platform." Meta spent $90 billion proving that wrong. This time, the consensus is "AI can replace everything," and every company is cutting headcount and funneling the savings into AI.

The only difference? Last time's consensus already broke. This time's is still holding.

But that's the thing about consensus. It doesn't erode slowly. It holds, holds, holds, then snaps. And the only thing that varies is how many billions get torched before the snap.

Meta isn't dumber than anyone else. It just bets bigger every time, so when the consensus flips, it hits the ground harder.

In 2021, the whole industry bet on the metaverse, and Meta changed its name. In 2026, the whole industry is betting on AI, and Meta is cutting a fifth of its workforce.

Five years from now, looking back, will this AI bet have been the right one?

Nobody knows.

 

If you find this helpful, feel free to follow us for future updates. ❤

• X : @blockflow_news
• Telegram: t.me/blockflownews