Trillion-Dollar Profits Now Require Fewer People Than Ever

Meta, Amazon, and Visa are laying off staff in 2026. Oracle laid off up to 30,000 employees via 6 AM email with no warning. Cloudflare cut 1,100 people—20 percent of its workforce. Block cut 40 percent of headcount and watched its stock jump more than 20 percent.

This is not a recession story. It is not a correction. The Big Five generated total revenues of $1.63 trillion in 2026, 81 percent more than five years earlier. These are the most profitable companies in the history of commerce, conducting mass layoffs anyway.

The 2026 tech labor market presents a paradox so clean it barely requires explanation. As of early August, there have been 322 layoff events impacting 205,832 workers, averaging approximately 949 job losses per day. Tech layoffs have hit 142,000 as profitable companies cut jobs to fund a combined $700 billion AI infrastructure buildout. Yet the defining feature of these cuts is not their scale. It is their simultaneity with record financial performance and record-breaking capital investment.

Amazon, Microsoft, Alphabet, and Meta have committed approximately $700 billion to capital expenditure in 2026—nearly double 2025 levels—most of it directed at AI compute, data centers, and networking infrastructure. S&P 500 companies spent $1 trillion on buybacks in 2025 while announcing 245,000 tech job cuts. The capital allocation strategy is explicit: redirect salary budgets toward AI infrastructure and share repurchases. This is not resource scarcity forcing hard choices. This is abundance creating different priorities.

The market has noticed. When Block cut 40 percent of its workforce, the stock jumped more than 20 percent. That signal travels instantly through the C-suite: CEOs get punished for headcount and rewarded for headcount reduction. The math is brutally simple. A mid-level engineer costs $200,000 to $300,000 annually all-in. That capital, redirected to an AI data center, becomes a depreciating asset that generates revenue without requiring healthcare, equity vesting, or annual raises. A person is a liability with upward wage pressure. Compute is a capital expenditure with predictable margins.

The industry will offer you an explanation. Fifty-four percent of layoff events this year explicitly cite AI, automation, or machine learning as the driving force, affecting 170,945 workers across 173 companies. The narrative is clean: technology is replacing workers, therefore companies must shrink headcount. Yet Sam Altman, CEO of OpenAI, has acknowledged that "some AI washing" is occurring where people blame AI for layoffs they would otherwise do. Translation: AI is the socially acceptable justification for what is actually a capital reallocation decision.

Consider the evidence. Oracle posted $13 billion in quarterly profit before laying off up to 30,000 employees via 6 AM email with no warning. That is not a company in distress; that is a company allocating capital according to a new preference function. Cloudflare cut 1,100 people while reporting record quarterly revenue of $639.8 million, up 34 percent year-over-year. These are not failing enterprises. These are enterprises that have calculated that the marginal return on human capital investment has declined below the marginal return on compute infrastructure.

The macroeconomic consequence is straightforward and grim. When the most profitable companies in the world stop requiring workers at the rate they once did, wage growth follows gravity. A labor market with 205,832 newly displaced workers is a labor market with downward pressure on compensation across the entire sector. Young engineers who expected 8 percent annual raises now negotiate for 3 percent. Mid-career managers compete for roles that weren't in the market six months ago. The entire wage structure adjusts downward.

This is not temporary. It is structural. Capital-intensive AI infrastructure represents a different economic model than labor-intensive software development. The capital has been committed. The trajectory is locked in. Meta, Amazon, Visa, and dozens of others are not laying off staff as a temporary measure pending recovery. They are laying off staff as a permanent recalibration of the business model. Profitability no longer requires people at the headcount levels that generated it.

You can call this efficiency. You can call this progress. What you cannot call it is neutral for wage growth. That story ended in 2026.