Meta is laying off humans to hire machines

On April 17, Reuters reported that Meta is targeting May 20 for the first wave of its 2026 layoffs, cutting roughly 8,000 employees, about 10% of its global workforce. More cuts are expected in the second half of the year. The scope could eventually reach 20% or more, depending on how the company's AI capabilities develop. At the same time, Meta's capital expenditure guidance for 2026 sits between $115 billion and $135 billion, nearly double the $72 billion it spent in 2025. Most of that money is going into data centers, GPUs, and cloud infrastructure to power its AI ambitions. The math is simple: headcount down, compute up. And Meta isn't alone. This is what AI-driven restructuring actually looks like when it stops being a thought experiment.

The numbers behind the pivot

Meta's workforce stood at roughly 79,000 employees as of late 2025. The first layoff wave alone would bring that down to around 71,000. If the rumored 20% total reduction materializes, the company could end 2026 with closer to 63,000 employees. Meanwhile, the spending is going in one clear direction. In April, Meta committed an additional $21 billion to CoreWeave for AI cloud infrastructure, on top of a prior $14.2 billion arrangement. The combined deal stretches through 2032. Meta is also expanding its in-house "Meta Superintelligence Labs" division, hiring researchers to develop what it calls a model with "godlike abilities." So the company is simultaneously getting smaller in people and bigger in compute. The signal to the market is unmistakable: Meta believes AI infrastructure generates more value per dollar than human labor in many of its current functions.

Everyone is doing this

Meta is the most visible example, but it's far from the only one. The pattern across Big Tech in 2026 is remarkably consistent. Microsoft cut 15,000 workers last year. Amazon laid off 30,000 in the past six months. Block, Jack Dorsey's payments company, slashed 40% of its workforce, more than 4,000 people, in a single day in February. Dorsey's shareholder letter was blunt: "Intelligence tools have changed what it means to build and run a company." Block's stock surged 22% on the announcement. Snap cut around 1,000 roles, citing AI-driven efficiencies. Even Crypto.com trimmed 12% of its workforce, with its CEO explaining that the eliminated positions were "roles that do not adapt in our new world." The numbers add up fast. According to industry tracking, nearly 80,000 tech workers were laid off in the first quarter of 2026 alone. About 48% of those cuts were explicitly attributed to AI and workflow automation. Collectively, the four major hyperscalers, Microsoft, Alphabet, Amazon, and Meta, are on track to spend somewhere between $650 billion and $700 billion on AI infrastructure this year. That's an increase of more than 60% from already historic levels in 2025.

The "AI-washing" question

Not everyone buys the narrative at face value. There's a growing skepticism about what analysts and commentators are calling "AI-washing" of layoffs, the practice of framing cost cuts as AI transformation when the reality might be more about margin optimization. Josh Bersin, a prominent HR industry analyst, pushed back on Block's announcement specifically. His argument: if AI is genuinely driving re-engineering, why wasn't it already happening incrementally? Great companies are always looking for ways to increase talent density. Cutting 40% of your workforce in one day looks less like a strategic AI transformation and more like a company that over-hired and is using AI as convenient framing. There's truth on both sides. Some of these cuts are genuine capability replacement. AI can now handle code review, content moderation, customer support triage, and many coordinator-level tasks that previously required dedicated headcount. But some of it is also opportunistic. When the market rewards layoffs with a 22% stock bump, executives have every incentive to frame their cost-cutting as forward-thinking AI strategy. The honest answer is probably that most companies are doing both at once: using real AI capabilities to justify overdue restructuring.

What the delta tells us

The most revealing data point isn't how many people are being cut. It's the contrast between the roles being eliminated and the roles being created. Meta is cutting across its operational workforce, but it's hiring AI researchers, infrastructure engineers, and data center specialists. The company is building out entire new divisions focused on superintelligence research. The eliminated roles tend to be in coordination, content operations, mid-level management, and functions where AI tooling has reached a tipping point of competence. This pattern holds across the industry. According to CompTIA, net tech employment in the U.S. is actually projected to grow by 1.9% in 2026, reaching about 9.8 million workers. The total number of open tech roles has been climbing since mid-2024. AI-specific roles are growing faster than any other category, and recruiter headcount is nearly back to its 2022 high. So the headline isn't "tech is shrinking." It's "tech is reshaping." The value is shifting from coordinative labor to infrastructure, from execution to judgment, from breadth of headcount to depth of capability. BCG's research supports this framing. Their microeconomic modeling suggests that over the next two to three years, 50% to 55% of U.S. jobs will be reshaped by AI, but most roles will remain in some form. Task automation doesn't straightforwardly equal job loss. It equals job transformation.

The scarring problem

The structural story is one thing. The human cost is another. Goldman Sachs published research in April drawing on 40 years of individual-level data to assess what happens to workers whose jobs are eliminated by technology. The findings are stark. Over the ten years following a job loss, real earnings for technology-displaced workers grow nearly 10 percentage points less than for workers who were never displaced, and 5 percentage points less than for workers displaced for other reasons. Technology displacement creates a specific kind of "scarring" that's harder to recover from than a normal layoff. When your entire skill category contracts, the job search takes longer, the new role often pays less, and the compounding effects on savings and retirement can stretch out for years. This hits a particular demographic hardest: workers in coordinative roles earning between $70,000 and $180,000, the middle managers, HR coordinators, analysts, and operations staff whose work is most susceptible to AI automation. These aren't entry-level workers who can pivot easily, and they're not senior executives with golden parachutes. They're the middle of the org chart, and the middle is getting compressed.

Where do the engineers go?

If Big Tech keeps cutting, where does all that displaced talent end up? Some will flow into startups. The combination of available talent and cheaper AI tooling is creating conditions for a new wave of lean, AI-native companies that can do more with fewer people. Every major layoff wave historically has seeded the next generation of startups, and this one should be no different. Some will go into government. The U.S. federal government lost around 20,000 tech workers last year and is now actively recruiting through programs like Tech Force. There's real demand for software engineers, cybersecurity specialists, and data analysts in the public sector, though the pay gap remains a significant barrier. Some will end up in industries that are expanding because of AI rather than contracting. Infrastructure companies, data labeling operations, AI safety firms, and cloud providers are all hiring. CoreWeave alone is scaling rapidly to handle Meta's billions in commitments. And some, inevitably, will face the scarring effects that Goldman described. Not everyone will land on their feet, and the transition period will be painful for many.

What this actually means

Meta's layoffs aren't surprising. They were inevitable the moment large language models got good enough at code review, content moderation, and operational coordination. The only question was timing, and now we have a date: May 20. What's worth paying attention to isn't the layoff itself, but the structural pattern it represents. Every major tech company is executing the same playbook: cut people, increase AI spend, reward remaining employees with higher talent density, and let the stock market applaud the efficiency gains. This is what "AI replaces jobs" looks like when it moves from conference stage speculation to quarterly earnings calls. It doesn't happen all at once. It happens in waves, company by company, function by function, with each round normalized by the last. The companies that come out ahead won't necessarily be the ones that cut the most. They'll be the ones that actually re-engineer their work around AI capabilities rather than just using the technology as cover for headcount reduction. The difference between genuine transformation and AI-washing will become apparent over the next 18 months, as the companies that cut deep either demonstrate real productivity gains or quietly start hiring again. For now, the direction is clear. Headcount down, compute up. The machines are being hired.