Google Just Dropped a Bombshell on AI Spending — Alphabet’s $200 Billion AI Bet Is Starting to Pay Off Big

Alphabet (GOOGL) has been spending aggressively on AI infrastructure, with capital expenditures projected to reach $195 billion to $205 billion by 2026. However, Google Cloud CEO Thomas Kurian says the company can expect a payback period of less than two years for its AI servers, and roughly one year for its proprietary silicon.

This suggests that Google can recycle capital into new infrastructure much faster than traditional capex would imply. Google's accelerator business, including Tensor Processing Units (TPUs), is already larger than the next-largest hyperscaler's accelerator business, and Google offers better price performance than alternatives. Revenue from Google Cloud grew 82% year-over-year in the second quarter, with a cloud backlog of $514 billion.

Nearly 90% of Fortune 100 companies are using Google's Gemini Enterprise. The company's AI spending looks less concerning if its infrastructure pays back quickly, and its custom silicon further improves economics. While not guarantees, Google's financial results suggest that the AI spending cycle could work out favorably. GOOGL stock trades at a reasonable valuation with a consensus "Strong Buy" rating.