The never-ending year of efficiency

TL;DR: Meta is reportedly planning its biggest layoffs since the 2022–2023 restructuring. Unlike last time, which was framed as a correction for pandemic overhiring, this round has everything to do with AI. If it happens, it makes Meta the latest in a slew of companies laying off workers while ramping up AI operations—and signals the efficiency push never really ended; AI just gave it a new cover.

What happened: Reuters reported on Friday that Meta is considering laying off around 20% of its nearly 79,000-person workforce—that’s a potential 15,000 jobs or more. It cited three anonymous sources, two of whom said top execs have already begun signaling the plans to senior leaders, asking them to start figuring out how to cut back. If the 20% figure holds, it would be the single largest round of cuts in the company's history—deeper as a share of headcount than either of the two rounds in 2022 and 2023, which totaled around 21,000 jobs across both. A Meta spokesperson called the report "speculative reporting about theoretical approaches," which is the corporate equivalent of not a denial.

Two forces, one outcome: There are two distinct drivers. The first is cost: Meta plans to nearly double its capex this year to as much as $135 billion, with the majority going to AI infrastructure—including a commitment to $600 billion on US AI efforts by 2028. The second is productivity: The cuts are apparently premised on AI tools enabling leaner teams to do more.

Meta’s AI play: The company has been spending aggressively on the other side of the ledger: acquiring agentic AI startup Manus for roughly $2 billion, buying the AI agent social platform Moltbook, and bringing in Alexandr Wang to lead its new Superintelligence Labs as part of a $14.3 billion investment in his company Scale AI—while reportedly offering some researchers pay packages of over $100 million to build out the team.

The company it keeps: This is part of a fast-moving pattern across Big Tech. Block cut 40% of its headcount in February, explicitly citing AI as the reason. Atlassian cut 10% of its workforce—about 1,600 people—earlier this month to fund AI investments and enterprise sales. Amazon eliminated around 16,000 roles in January, citing the need to reduce layers and bureaucracy as it ramps up AI spending.

Bottom line: Mark Zuckerberg called 2023 the "Year of Efficiency"—a name that implies a beginning and an end. What this round of potential cuts makes clear is that efficiency seemingly never ended. The new equation is: Cut labor to fund compute, then bet the compute pays for itself. —SM