Meta beats revenue estimates, ups long‑term capex and signals AI-driven talent/monitoring shifts

Meta just showed the split screen of the AI boom. On one side, the business is throwing off huge amounts of cash. On the other, the company is spending even more aggressively to build the infrastructure it thinks the next phase of AI will require. That tension got sharper on April 29, when Meta posted a strong first quarter, raised its capital spending outlook, and kept talking like an AI lab with an ad machine attached. (investor.atmeta.com) ### What happened in the quarter? Revenue came in at $56.31 billion for the March quarter, up 33% from a year earlier. Operating income was $22.87 billion, and operating margin held at 41%. Daily active people averaged 3.56 billion in March. That was still up 4% year over year, even though Meta said internet disruptions in Iran and a WhatsApp restriction in Russia hurt usage sequentially. (investor.atmeta.com) ### Why did investors still flinch? Because the bigger story was spending. Meta raised its 2026 capital expenditure forecast to $125 billion to $145 billion, from a prior $115 billion to $135 billion range. The company said the increase reflects higher component pricing this year and extra data-center costs needed to support(investor.atmeta.com)ills. (cnbc.com) ### What is Meta actually buying? Compute — lots of it. Q1 capital expenditures were $19.84 billion, and Zuckerberg framed the quarter around “Meta Superintelligence Labs” and a push to deliver “personal superintelligence.” Strip away the branding and the point is simple: Meta thinks better models, more inference, and more AI features across its apps will keep lifting the ad business now, while setting up new products later. (investor.atmeta.com) ### Why does the workforce angle matter? Because Meta is not just buying GPUs. It is reorganizing work around AI at the same time. CNBC reported last week that Meta said it was cutting 10% of its workforce, part of a broader pattern where the companies spending the most on AI infrastructure are also trimming headcount and leaning harder on automation. That makes the earnings story less about one quarter’s beat and more about a new operating model. (cnbc.com) ### What’s going on with employee monitoring? Reuters reported that Meta is installing software on U.S. employees’ computers to capture mouse movements, clicks, keystrokes, and occasional screen snapshots. The tool — Model Capability Initiative, inside a broader “AI for Work” effort now called Agent Transformation Accelerator — is meant to generate traini(cnbc.com)formance reviews. (tech.yahoo.com) ### Why is that such a big signal? Because it shows how far Meta wants to push “agentic” work. Bosworth told employees the goal is a future where agents “primarily do the work” and humans direct, review, and improve them. That is a much more radical vision than just giving workers a chatbot. It implies Meta wants software that can actually operate workplace tools, not merely answer questions about them. (tech.yahoo.com) ### Are workers even good at using AI yet? Not consistently. Former Meta engineering manager Kun Chen said many CTOs are seeing only 10% to 15% productivity gains from AI because most employees use it in a shallow way. His sharper point was that maybe 2% of engineers have figured out how to use AI “very effectively,” and those people are getting the mos(tech.yahoo.com)tech. (tech.yahoo.com) ### So what’s the real takeaway? Meta’s quarter says the ad engine is strong enough to fund an enormous AI land grab. But the catch is that this is changing the company from both ends at once — more capex at the top, more pressure on labor in the middle, and more software watching how work gets done so future software can do more of it. (investor.atmeta.com)efault.aspx))