AI Exchanges: How tech giants are navigating the AI landscape

Investment Cycle and Capital Expenditure Trends

• Major technology companies have reaffirmed commitment to capital expenditure (CapEx) through the end of 2025 to support AI workloads, despite macroeconomic volatility.
• The current period represents the third year of an investment cycle, now at peak capital intensity, with meta approaching 40% CapEx-to-revenue ratios.
• Goldman Sachs projects CapEx growth to slow to mid-teens in 2026, down from 40-60% growth rates observed in 2024.
• Specific 2025 CapEx forecasts include approximately $75 billion for Alphabet, nearly $70 billion for Meta, and $100–$110 billion for Amazon (collectively exceeding $250 billion).
• Unlike previous cycles, incumbent giants are deploying massive capital both offensively and defensively to prevent disruption, a strategy made possible by their unprecedented balance sheet scale.
• Tariffs are already impacting CapEx budgets as embedded costs for imported hardware and data center materials, effectively increasing the cost of goods sold without necessarily reducing capacity plans.

Market Positioning and Phases of AI Evolution

• Eric Sheridan identifies the current era as the early innings of a "third computing shift" (Web 3.0), transitioning from the infrastructure layer (large language model training) to the platform layer, with the application layer to follow.
• Amazon CEO Andy Jassy characterized the AI wave as "first batter, second strike of the first inning," signaling that the industry is in the novelty phase with significant territory remaining.
• While the cycle has accelerated rapidly—ChatGPT reached 800 million monthly active users within two and a half years of launch—the application layer for both consumer and enterprise use remains the primary uncertainty for future returns.
• Investors are increasingly focused on the "arms dealers" (hyperscalers and infrastructure providers) rather than the potential consumer application winners, though the latter remains the "to be determined" component.
• Cloud revenue growth rates (AWS at 17%, Google Cloud at 28%, Microsoft Azure in the 30s) are driving current investor attention due to direct correlations between CapEx and cloud workload monetization.

Search, Monetization, and Consumer Behavior

• Contrary to fears of cannibalization, search volumes have not declined; instead, the total number of queries to computers has exploded, with commercial monetization remaining predominantly with Google.
• Alphabet is actively adapting its product suite by integrating AI Overviews, Gemini desktop interfaces, and standalone apps to maintain relevance within the search ecosystem.
• Meta and Alphabet are realizing immediate internal efficiency gains and ROI from AI, particularly through the automation of ad creation, placement, and measurement, which improves return on ad spend.
• Consumer habits regarding technology adoption are projected to shift slowly, taking a decade or more to fully realize, contrasting with the rapid pace of innovation observed among early adopters.
• Amazon is pivoting its consumer strategy by promoting its Alexa app as a primary AI assistant, competing with standalone apps from Meta, Google, and private firms like Claude and Perplexity.

Investor Sentiment and Valuation Metrics

• Technology giants involved in the AI shift are trading at valuations that are historically conservative, with Alphabet trading below market multiples in the forward year and Meta trading only slightly above.
• This stands in contrast to the 2000 tech bubble era and prior shifts where incumbents were often disrupted by new entrants; today's incumbents have the capital scale to sustain the transition.
• Investor patience remains low, with narratives susceptible to rapid reversal if value realization does not occur within a single earnings cycle.
• Market sentiment is currently bifurcated into "pre-tariff" and "post-tariff" narratives, with the latter dominating recent conversations regarding Amazon and broader economic risks.
• Goldman Sachs analysts note that while the infrastructure layer has peaked in rate of change, the market is currently awaiting proof points from the application layer to justify sustained high-level CapEx beyond 2026.