The productivity gains are real. The record profits are real. The thousand people who lost their tech jobs today are also real.
Analysis tracked by TechStartups shows tech companies are laying off approximately 1,000 workers per day in 2026, with AI automation increasingly cited as the driving factor. The pattern is consistent and uncomfortable: a company announces record earnings in the same quarter it announces significant workforce reductions, crediting AI tools that now handle code generation, customer support, data analysis, and document processing without human intervention. From a shareholder perspective, the math is unambiguous. From a workforce perspective, it is something else entirely.
Meta's situation illustrates the dynamics clearly. The company's $14 billion investment in AI — including the acquisition of Scale AI founder Alexandr Wang's team and the launch of Meta Superintelligence Labs — is happening alongside relentless pressure to prove AI can generate revenue beyond advertising, which still accounts for 98% of Meta's income. Big Tech has promised investors an AI future while funding it by cutting costs today. Workforce reduction is the fastest lever available. The emerging categories of AI work — model trainers, infrastructure engineers, evaluation specialists — take years to build up to the scale of what is being shed.
The honest version of this story is that labor market disruption from AI is front-loaded. The jobs being eliminated are visible and immediate; the jobs being created are diffuse, harder to qualify for, and slower to materialize. That gap is landing hardest on mid-level knowledge workers who spent a decade being told that creative, analytical, communication-heavy work was automation-proof. It turns out "automation-proof" meant "until the next generation of models."
None of this makes AI bad, or even net-negative for humanity in the long run. But "in the long run" is cold comfort if you are one of the thousand today. The central question for 2026 is not whether AI will eliminate jobs — it clearly is. The question is whether the productivity gains get broadly distributed or continue to accrue almost entirely to people who already had all the money. History is not encouraging on this front. We are still writing this one.
Source: TechStartups