Meta is laying off 10 percent of its staff

The freeze on 6,000 open positions may matter more than the layoffs themselves: Meta isn't just trimming headcount, it's signaling that the hiring pipeline built for broad product expansion is being redirected toward a narrower AI-first org. The people leaving aren't necessarily the ones building models.

This fits a pattern Modelwire has been tracking across the major labs. OpenAI's April departures, covered across TechCrunch and The Verge around April 17, showed a company dissolving teams (Sora, the science unit) that no longer mapped to its enterprise-focused priorities. Meta's move is structurally similar: capital that was spread across consumer products and general headcount is being concentrated. The difference is scale. OpenAI shed specific project teams and executives; Meta is cutting roughly 8,000 roles while simultaneously pouring money into infrastructure. That combination, fewer people plus more compute spend, is the clearest statement yet about where these companies think value actually accrues.

Watch whether Meta's Q2 earnings call in late July shows AI infrastructure capex continuing to rise even as total headcount falls. If both trend in opposite directions simultaneously, that confirms the thesis that labor is being substituted for compute, not just reduced for efficiency.