META Commits $145B Capex While Cutting 8,000 Jobs in a Wholesale Pivot to AI

Meta's restructuring is far more aggressive than traditional cost-cutting. The company is simultaneously culling 8,000 roles, canceling 6,000 open positions, and reallocating 7,000 employees into AI-centric units while projecting $145B in capex spending. This is not austerity; it is a wholesale pivot toward artificial intelligence infrastructure and compute at the expense of organizational overhead.

The capex guidanceCompany-issued forecasts of future financial performance. is staggering. At $145B, Meta's spending trajectory rivals that of major semiconductor manufacturers and exceeds most energy companies. The company is building custom silicon, data center infrastructure, and foundation model capabilities at scale. This spending will flow through to NVIDIA, Broadcom, and other chip suppliers, further validating the AI capex supercycle thesis. Meta is betting that proprietary AI models, trained internally and deployed across its social platforms, will drive future user engagement and advertising yield.

The workforce restructuring reveals meta's conviction that traditional software engineering and content moderation roles will be increasingly automated or outsourced. By consolidating technical talent into AI research and infrastructure teams, Meta is essentially hedging against the commoditization of application-layer software. The company is doubling down on first-mover advantage in large language models and multimodal systems.

The risk is execution. Capital spending commitments at this scale require high utilization and ROI within a reasonable timeframe. If Meta's internal AI models fail to outperform open-source alternatives, or if customer demand for AI-driven features underperforms expectations, the company faces a valuation reset. Additionally, the labor market impact of laying off 8,000 engineers signals that traditional software roles are becoming less valuable, which could depress wage expectations across the tech sector and accelerate automation pressure.