Gene Munster Put 70% of His Money In AI —What He's Buying Now

Gene Munster's Deepwater Asset Management recently sold half its Nvidia stake — a signal that even infrastructure bulls are repositioning. Munster still owns the stock and considers infrastructure «significant,» particularly as inference computing (the thinking part of AI) scales to become hundreds of thousands of times larger than training workloads. Nvidia will benefit, but the position is now «smaller» and «material» rather than concentrated.

The calculus reflects a nuanced view: if we're truly in the second inning of AI, the basket approach that worked in 2023–2024 is giving way to selectivity. Munster argues investors can fractionally outperform the Nasdaq by owning selective big tech, but «the real upside is going to come from these smaller companies, these companies that are 500 billion in enterprise value and less.» He's not abandoning mega-cap — he's rebalancing toward where exponential growth remains possible.

Munster also owns Amazon for its AWS exposure and «compelling valuation,» but pointedly avoids Microsoft. The distinction: Amazon benefits from inference infrastructure, while Microsoft's seat-based software model faces existential questions. Munster puts a 35% probability on AI coding and agents seriously disrupting companies that charge per human seat, as one license could soon control hundreds of AI workers.

Munster recently added to his Apple position based on a bet that personalized AI will transform the company's trajectory over five years. The catalyst: a major Siri overhaul codenamed «Apple Intelligence,» expected in June, that aims to compete directly with ChatGPT and Gemini while leveraging personal context and privacy. Bloomberg's Mark Gurman has reported specifics suggesting Siri will become a true AI assistant, not just a voice interface.

The bar is high. Apple «has not shown that they have competency in AI,» Munster admits, and there's «not a must-have» feature in Apple Intelligence today. If the June launch falls flat, the stock could drop 10–15%. But Munster believes investors will grant Apple extended patience — «as much time until some other competitive AI device really captures the imagination of investors» — because Apple and Google are the only two companies positioned to solve personalized, privacy-preserving AI at scale.

Munster's conviction is striking: in rapid-fire questions, he chose Apple over Alphabet, picked AI Siri over ChatGPT, and named Apple as both the stock that «hasn't had its moment yet» and the one AI stock he'd hold for the next 12 months. The thesis: if Apple delivers, the rerating will be dramatic. If not, the market will wait for the next attempt.