2 Magnificent Artificial Intelligence (AI) Stocks to Buy and Hold for the Next 20 Years

The article argues that Amazon and Microsoft are positioned to benefit most from AI infrastructure spending over the next 20 years. Amazon’s AWS grew 28% in Q1, its fastest pace in 15 quarters, while its custom AI chip business rose at a triple-digit rate and the company is investing $200 billion in infrastructure this year. Microsoft’s AI-related business has reached $37 billion and is growing 123%, while Azure revenue rose 40% in the latest quarter.

Analysis

The market is still underestimating how much of AI’s early profit pool will accrue to infrastructure owners rather than model builders. For AMZN and MSFT, the key second-order effect is that AI workload growth turns their cloud businesses into higher-throughput toll roads: even if model differentiation commoditizes, token consumption and inference demand should extend the monetization runway for years. That creates a structurally better earnings mix because incremental AI revenue is disproportionately attached to existing fixed-cost platforms, which should amplify operating leverage once capex intensity normalizes.

The bigger implication is competitive entrenchment. Amazon can subsidize AI infrastructure with commerce cash flow, while Microsoft can cross-sell AI into its installed productivity base, so both can outspend smaller peers without needing immediate payback. That is bad news for mid-tier cloud providers and enterprise software vendors lacking either scale or distribution; they face rising customer expectations for AI-native features while competing against bundled pricing from hyperscalers. Expect the displacement to show up first in renewal cycles over the next 6-18 months, not instantly in headline revenue.

The main risk is that the current optimism assumes AI demand stays linear while capex is lumpy. If enterprise inference spend slows or pricing falls faster than utilization rises, the market could punish both names for a few quarters because depreciation and amortization will lag the revenue ramp. The contrarian read is that consensus is focused too much on total AI demand and not enough on power, networking, and chip supply bottlenecks; those constraints could delay monetization and create tactical volatility even in a multi-year bullish thesis.

Key Decisions for Investors

• Maintain core longs in AMZN and MSFT over a 12-24 month horizon; use any 5-8% post-earnings drawdown to add, as the market may overreact to capex optics before free-cash-flow inflects.
• Pair trade: long AMZN/MSFT vs short a basket of second-tier cloud/software names with weaker distribution moats; the setup favors hyperscalers capturing AI spend while smaller vendors pay for integrations.
• Consider buying 6-12 month call spreads on AMZN or MSFT instead of outright calls to express upside from AI monetization while limiting multiple-compression risk if capex proves too heavy.
• Avoid chasing NVDA purely on this thesis; the article supports durable demand, but the best risk/reward appears in the platform owners where AI spend converts into recurring service revenue.