The AI Stock Trap David Booth Just Warned You About

The AI Stock Trap David Booth Just Warned You About

David Booth, the founder of Dimensional Fund Advisors, just said something that should make every AI stock picker pause: picking winners in artificial intelligence can backfire. He was talking about Nvidia, the chip company that already commands so much of the AI infrastructure market that beating it feels almost impossible.

The problem is not complicated. Investors have convinced themselves that because AI is here, one of three things must happen: Nvidia keeps dominating forever, or a new winner emerges, or the whole cycle crashes. They then bet accordingly. What they forget is that past performance tells you almost nothing about which outcome wins.

Why Nvidia's Lead Is Deceptive

Nvidia made a $267 billion warning this month about potential crashes in Micron stock before 2029 ends. That single statement shows how much power Nvidia has: it can move other companies' valuations just by talking. The company controls the GPU architecture that trains large language models. It owns the software ecosystem. It has the economic moat. But Booth's point is sharper: that dominance makes it almost impossible to find a better alternative, which means the stock is priced for perfection.

When you price a stock for perfection, any stumble becomes a catastrophe. Competition from AMD or Intel, a slowdown in enterprise AI spending, or simply profit-taking can obliterate years of gains in weeks. The very thing that makes Nvidia look safe (its market share) also makes it fragile (it has nowhere to grow into).

The Diversification Mistake

Look at the headlines this week. Anthropic is signing a $35 billion computing deal with Lambda. Apple is trying to prove it is more than just an iPhone company under John Ternus. Tesla climbed on fresh investor optimism. Mediatek just signed a uniquely important deal with Nvidia itself.

Each of these stories made someone feel smart about their bet. But that feeling is exactly the trap Booth is warning about. You cannot know which company will matter most in 2027 or 2028. You can know that AI infrastructure will grow. You can know that semiconductor demand will expand. But betting on one horse in that race, or even three, leaves you vulnerable to being wrong about which horse wins.

What to Do Instead

The practical lesson here is not to avoid AI stocks. It is to own them as a sector, not as a guessing game. When Exxon is not invited to oil meetings and pipeline stocks are making $4.4 billion acquisitions to keep dividend yields stable, investors get nervous and reach for growth. AI feels like growth. But AI is infrastructure now. Infrastructure is not a bet on the right company. It is a bet on the whole category.

You can own semiconductor ETFs that hold Nvidia, AMD, Micron, and others without pretending you know which one will win. You can own software companies building on top of those chips without betting your portfolio on Anthropic over others. The winners will emerge, but they will emerge for reasons you cannot predict today.

Booth's warning is really saying: if you have to be that smart to make money, you are not investing. You are gambling. The stock market has room for both, but they are not the same thing.

The bottom line

AI is not a stock pick. It is a sector. Booth knows that betting on the right company is harder than it looks because once a company dominates (like Nvidia does now), it has already won and lost room to run. Own the infrastructure category broadly and sleep better.

This is educational information, not financial advice.