AI Worries Erase $1T from Big Tech

- The planned 2026 capital expenditures for AI infrastructure are substantial, with Amazon earmarking around $200 billion, Google's parent Alphabet planning for $175-$185 billion, and Meta intending to spend between $115-$135 billion. - Individual market capitalization losses in early 2026 have been significant, with Microsoft's value dropping by approximately $613 billion, Amazon's by $343 billion, Apple's by $256 billion, and Nvidia's by nearly $90 billion. - Investor reaction has not been uniform; Meta's stock rose after its spending announcement due to strong AI-driven advertising revenue, while Amazon's shares fell nearly 10% following its capex forecast. - Analysts have noted a clear shift in market sentiment, with Jesse Cohen, senior analyst at Investing.com, observing a "growing divide between tech companies' AI ambitions and Wall Street's patience for open-ended investment cycles." - The massive spending has led to concerns about the depreciation lifecycle of specialized hardware, with some analysts suggesting that the economic obsolescence of GPUs in 2-3 years is much faster than the 5-6 year depreciation schedules used by some companies. - In contrast to the tech giants investing heavily in their own AI infrastructure, companies in the semiconductor sector have seen their market values rise. TSMC and Samsung Electronics, for example, have experienced gains as they are key suppliers for the AI hardware buildout. - The concern over profitability is not just about the cost of infrastructure, but also the potential for new AI models to disrupt existing software-as-a-service (SaaS) business models, a fear that has contributed to sell-offs in the software sector. - The scale of the projected AI-related spending by just four major tech companies in 2026, estimated to be around $660 billion, exceeds the GDP of many countries, including Singapore and the United Arab Emirates.