Meta Fired 8,000 People to Hire One 28-Year-Old

$135 billion. That is the amount Meta plans to spend on artificial intelligence infrastructure in 2026 alone — roughly the GDP of Hungary. To fund it, the company is laying off 8,000 people on a single day: May 20. The architect of this transformation is not a grizzled Silicon Valley veteran but a 28-year-old former startup founder who convinced Mark Zuckerberg that the path to superintelligence runs through his lab.

The Biggest AI Bet in Corporate History

From $72B to $135B in One Year

Meta’s 2026 capital expenditure guidance of $115 billion to $135 billion represents an increase of roughly $60 billion over the $72.2 billion spent in 2025. Nearly all of it is flowing into data centers, Nvidia GPUs, custom silicon, and the infrastructure backbone for its Llama model family and the newly created Superintelligence Labs.

Wall Street, remarkably, is cheering. Meta shares surged 10% when the spending plan was first announced in January, and analysts have broadly maintained buy ratings. The thesis is straightforward: if AI delivers even a fraction of its promised productivity gains across Meta’s 3.3 billion daily active users, the return on investment could dwarf the capital outlay.

The 28-Year-Old Running Meta’s AI Future

Alexandr Wang and the Scale AI Acquisition

Alexandr Wang founded Scale AI at age 19 and built it into one of the most critical data infrastructure companies in the AI ecosystem. In June 2025, Meta acquired a 49% stake in Scale AI for $14.3 billion and installed Wang as Chief AI Officer, placing him at the helm of a newly created division: Meta Superintelligence Labs.

Wang’s appointment marked the effective end of an era. Yann LeCun, Meta’s legendary Chief AI Scientist who had shaped the company’s AI research direction for over a decade, departed in late 2025, citing fundamental disagreements over research direction. Meta subsequently cut 600 researchers from its FAIR (Fundamental AI Research) lab and consolidated AI efforts under Wang’s leadership.

Muse Spark: First Blood

The first tangible output of Wang’s lab arrived earlier this month: Muse Spark, a frontier large language model that reportedly delivers competitive performance on multimodal perception, reasoning, and agentic tasks — at a fraction of the compute cost of Meta’s older Llama 4 mid-size variant. While independent benchmarks are still emerging, the model represents a philosophical shift from Meta’s previous open-source-first approach toward a more vertically integrated strategy.

8,000 Jobs Gone, New Titles Created

The May 20 Layoff and Organizational Overhaul

On April 23, Meta confirmed it will lay off approximately 8,000 employees — 10% of its 78,865-person workforce — with the cuts taking effect on May 20. An additional 6,000 open positions will be eliminated. The company has signaled that further reductions are planned for the second half of 2026.

But this is not simply a cost-cutting exercise. Maher Saba, who now heads the new Applied AI Engineering division under CTO Andrew Bosworth, circulated an internal memo on April 14 outlining a sweeping reorganization. Traditional engineering and product roles are being replaced with AI-native titles: “AI builder,” “AI pod lead,” and “AI org lead.” Roughly 1,000 employees have already been reassigned under the new structure, with engineers from across the company being funneled into the Applied AI organization.

The Two-Team Model

The Applied AI Engineering division is split into two teams: one focused on interfaces and tooling (the surfaces through which users and developers interact with AI), and another on task execution, data generation, and evaluations (the backend muscle that powers AI agents). This structure mirrors the architecture of modern AI systems themselves — a clean separation between the orchestration layer and the execution layer.

What This Means for the Industry

Meta’s restructuring is the most aggressive example yet of a phenomenon sweeping Big Tech: the conversion of human payroll into AI capital expenditure. Microsoft has made similar moves, and across the industry, companies are learning that the marginal cost of an AI agent is orders of magnitude lower than the marginal cost of a human employee — at least for routine cognitive tasks.

The question is whether this bet pays off. Meta is essentially wagering that Alexandr Wang can build a superintelligence lab capable of producing models that justify $135 billion in annual infrastructure spending. If Muse Spark is any indication, the early returns are promising. But the gap between a promising first model and a self-sustaining AI ecosystem is vast, and the competition from OpenAI, Google DeepMind, and Anthropic is relentless.

For the 8,000 workers losing their jobs on May 20, the strategic logic is cold comfort. But for the AI industry as a whole, Meta’s move marks an inflection point: the moment when the world’s largest social media company decided its future belongs not to the people who built it, but to the machines they are building.

Related

• OpenAI Just Made Every Other AI Look Outdated
• This $0.14 Model Just Embarrassed OpenAI and Google
• They Built One AI Engineer — Now They’re Worth $25 Billion
• One Company Just Attracted 90 Trillion Won in Investment

Sources

1. Axios — Meta to lay off 8000 as part of AI efficiency push (April 23, 2026)
2. Variety — Meta Laying Off 8,000 Employees Amid Surge in AI Spending
3. CNN — Meta to cut 10% of staff as it pours billions into AI
4. The Next Web — Meta to cut 8,000 jobs on 20 May with more layoffs planned