How AI is Draining Google’s Finances

Summary

Alphabet’s third-quarter results unsettled investors as AI-related spending rose sharply while revenue from its historically dominant, highly profitable Search business fell short of expectations. Shares dropped about 10% in after-hours trading, erasing roughly $400 billion in market value.

Spending and AI strategy

• Costs increased: Alphabet spent $87 billion during the quarter, up 22% year over year. Spending for the first three quarters totaled $236 billion, up 21%.
• AI investment has two main components:
  • Capital expenditure: The company spent $51.4 billion on capital assets during the quarter. CFO Anat Ashkenazi said roughly 65% went to computing equipment—including GPUs, CPUs, TPUs, and networking—and 35% to data-center infrastructure, such as buildings, power, and cooling.
  • TPU inventory: Inventory increased by $10.9 billion year over year, reportedly mostly from building Google’s specialized AI chips. Google says it manufactures components and builds TPU inventory after signing customer contracts, which could reduce the risk of producing chips without buyers.
• Two-sided AI strategy: Google is building computing capacity to support its own AI services while also developing and selling TPUs to customers including Anthropic, Apple, and Oracle. The strategy requires substantial upfront investment, and its returns depend on customer demand and successful execution.

Revenue mix and business risks

• Total revenue: $127.6 billion, up 25% year over year, just below analysts’ forecast of $128 billion.
• Google Services: $96 billion, up 11% year over year, but below expectations.
  • Google Search: $63.3 billion, about $1.5 billion below forecast and roughly flat from the previous quarter.
  • Alphabet acknowledged that AI-driven search queries, including those using AI Mode, may generate less revenue because they are monetized at lower rates.
• Google Cloud: About $30 billion, roughly $2 billion above forecast. This offset weakness in Services and helped bring total revenue close to expectations.
• The market’s concern is not just the revenue shortfall. Search has long been Alphabet’s most reliable profit engine, while investors may not yet see Cloud or AI as equally dependable or protected by comparable competitive advantages.

Cash flow and investor concerns

• Free cash flow was negative $9 billion, compared with negative $5.9 billion in the previous quarter.
• The video describes this as only the second quarterly loss since Alphabet went public in 2004.
• Alphabet faces a central business tension: it is investing heavily in AI while its traditional Search model faces potential disruption—and AI queries currently appear less lucrative. Strong Cloud growth and contracted TPU demand offer possible paths to returns, but the payoff remains uncertain.

Frameworks and business takeaways

• Separate AI spending into infrastructure and inventory: Evaluate data-center and computing investments separately from chips built for sale, since they have different cost structures and routes to revenue.
• Use customer commitments to guide capacity and inventory: Google says it builds TPU inventory after contracts are signed. This links production to demand, though delivery and profitability still depend on execution.
• Track revenue by business segment, not just in aggregate: Cloud growth brought total revenue close to forecast, but did not erase concerns about weakness in the more profitable Services business.
• Monitor monetization alongside AI usage: More AI activity does not automatically mean more revenue if AI queries generate less advertising income than traditional searches.

Presenters and sources

The video is narrated by an unnamed TLDR Business presenter. It cites Alphabet’s quarterly results and earnings call, including comments from CFO Anat Ashkenazi and Google Cloud CEO Thomas Kurian.