Meta has become two businesses at once, somewhere between the jubilant tone of an earnings call and the subdued dread of a Thursday afternoon all-hands. One is the version that Mark Zuckerberg frequently discusses on stage, one that is illuminated by advances in artificial intelligence and optimistic forecasts regarding productivity. The other is the one being typed about, anonymously, late at night on Blind, where the words “dead and depressing” now sit at the top of a thread about a company that, not long ago, employees fought to get into.
A portion of the story is revealed by the numbers. Meta will cut roughly 8,000 jobs next month, about one in ten employees, and quietly close some 6,000 open roles it had been hiring for. Spending on AI this year is expected to hit $135 billion, a figure that, according to one person who saw the internal memo, nearly equals what the company spent on AI in the previous three years combined. That is an incredible reallocation of capital, and it is occurring quickly.
| Information | Details |
|---|---|
| Company | Meta Platforms, Inc. |
| CEO | Mark Zuckerberg |
| Headquarters | Menlo Park, California |
| Layoffs Announced | Approximately 8,000 employees (10% of workforce) |
| Effective Date | May 20, 2026 |
| AI Spending in 2026 | $135 billion |
| Open Roles Closed | Around 6,000 |
| Negative Employee Sentiment on Blind | Jumped from 20% (2024) to 83% (2025–26) |
| Previous Layoff Rounds in 2026 | Two smaller rounds, ~2,000 workers |
| Source of Sentiment Data | Blind anonymous workplace platform |
The mood is more difficult to chart. According to data shared with Fast Company, negative posts about AI at Meta on Blind have climbed to 83% this year, up from around 20% in 2024. It’s not a gradual drift. It’s a collapse. In April alone, more than 500 posts on the platform discussed the layoff announcement. One worker, asked about the changes, told the BBC the company had become “obsessed with AI.” The word choice matters. Obsession implies something has tipped past strategy into something less rational, less examined.
This story has a certain detail that is difficult to forget. According to reports, Meta told staff members that it would start monitoring their mouse movements, clicks, and keystrokes on work computers the same week the layoffs were announced. This information would then be fed into AI training data. One worker described it as “dystopian.” Even from the company’s point of view, it’s difficult to ignore the reasoning: in a way, the employees being replaced are also being asked to train their replacements. It’s unclear if anyone at Meta stopped to consider the timing’s optics.
Zuckerberg has been remarkably forthright. During the January earnings call, he stated that tasks that previously required large teams could now be completed by “a single very talented person.” A few weeks later, while seated next to Satya Nadella, he projected that perhaps half of Meta’s development work would be completed by AI instead of humans in a year. Investors appear to have faith in him. Employees, increasingly, do not — at least not in the optimistic sense executives mean.
Of course, this isn’t just a Meta story. On the same day, Microsoft gave about 7% of its American employees the option to retire voluntarily. Amazon has cut at least 30,000 workers in six months. In March, Block cut its workforce in half. Even if the long-term payoff isn’t obvious, the pattern is. Walking through the coverage, it seems as though the industry is placing a huge wager that even its supporters are still unable to fully explain.
After reading the memos, the Blind threads, and the well-crafted statements, one question remains that no one at Meta seems eager to address. What precisely are the remaining 90% of workers being asked to develop, and for whom, if AI is truly handling the tasks of entire teams?
