Billionaire David Tepper of Appaloosa Is Piling Into the AI Revolution With These 3 Big Purchases

David Tepper's Appaloosa materially increased AI-linked exposure in Q1, lifting Amazon stake 98% with 2,140,609 shares, buying 4,482,720 Uber shares for a 242% increase, and initiating a new 281,250-share position in SanDisk. The article argues all three benefit from AI adoption: AWS and ad growth at Amazon, AI-driven routing and pricing at Uber, and surging NAND demand for AI data centers at SanDisk. This is constructive for the names but is primarily a bullish commentary on billionaire positioning rather than new company-reported fundamentals.

Analysis

The common thread is not “AI exposure” but pricing power embedded in distribution bottlenecks. AMZN and UBER are the cleaner beneficiaries because their AI spend is mostly efficiency-enhancing while demand is already scaled; that creates a double lever on margins if routing, ad targeting, and fulfillment utilization keep improving. SNDK is different: it is a supply-constrained beneficiary where earnings are likely being pulled forward by inventory tightness and capex urgency, which can sustain outsized revisions for several quarters but also makes the setup fragile if NAND capacity catches up faster than expected.

The second-order winner is the hardware ecosystem around AI data centers and the losers are the companies without either proprietary demand or scarce components. If memory/storage remains tight, hyperscaler procurement will stay disciplined on adjacent spend, which can pressure lower-value infrastructure vendors and delay some noncritical capex. For AMZN specifically, stronger AWS AI monetization can spill into advertising and logistics utilization, while UBER’s AI stack lowers unit cost per trip and increases take-rate durability rather than just top-line growth.

The market is likely underpricing how much of this is duration, not just near-term earnings. SNDK has the most reflexive upside but also the highest reversal risk: once consensus fully reflates, the stock becomes vulnerable to any sign of customer inventory normalization or pricing peak, likely on a 3-6 month horizon. AMZN and UBER are lower-beta ways to express the same AI-theme with better fundamental compounding over 12-24 months.

The contrarian miss is that this is not a pure multiple-expansion story for all three names. SNDK may already be discounting a very aggressive earnings path, so chasing it outright is less attractive than owning it with defined downside. By contrast, AMZN and UBER still have room for estimate revisions to drive the next leg higher, especially if investors continue to treat them as consumer/transport names rather than AI productivity beneficiaries.

Key Decisions for Investors

• Long AMZN on 3-6 month horizon vs. the NASDAQ: use it as the highest-quality AI compounder with operating leverage from AWS and ads; favorable if estimates keep grinding higher and downside is muted relative to AI hardware.
• Long UBER vs. LYFT-style exposure or unprofitable mobility names over 6-12 months: the thesis is AI-driven efficiency plus market-share durability, with better cash-flow conversion and less multiple risk than pure growth comps.
• Trade SNDK with a call spread rather than outright stock: buy 3-6 month call spreads to capture continued estimate revisions while capping damage if NAND pricing normalizes faster than consensus expects.
• Pair trade: long AMZN / short a basket of low-quality AI infrastructure or non-differentiated tech laggards over 6-12 months; the market should keep rewarding companies that monetize AI through distribution, not just those that mention it.
• Set a tactical profit-taking rule on SNDK after the next two earnings prints: if management commentary shifts from scarcity to capacity expansion, reduce exposure quickly because the stock’s re-rating can unwind sharply.