Just came across something interesting about how the smartest money in the world is actually positioning itself around AI, and it's not quite what you'd expect.

Warren Buffett has over $75 billion tied up in just three stocks, and while most people assume he's chasing AI hype, the reality is way more nuanced. His actual thesis seems to be about finding the best way to invest in AI through companies with unbreakable competitive advantages.

Let's break this down. Apple is his biggest position at $67.44 billion. Most assume it's because of their AI features, but that's missing the point entirely. Buffett's been holding Apple since 2016 because of the insane customer loyalty, the management team, and honestly, their buyback program is absolutely brutal - they've spent over $816 billion retiring shares since 2013. That's the kind of financial discipline he loves. Now they're layering in Apple Intelligence across products, which is interesting, though he's actually been trimming the position lately. The valuation got stretched.

Then there's Alphabet. The 17.8 million Class A shares Berkshire picked up are worth about $5.62 billion. Google basically owns search with 89-93% market share globally. That's the moat. But here's what people miss - their cloud division is where the real AI action is happening. Google Cloud is growing over 30% year-over-year by integrating generative AI and LLM capabilities for enterprise clients. During Q3 when Buffett was buying, the stock was trading at a forward P/E of 16-22. That looked cheap for that growth rate.

Amazon rounds it out with $2.34 billion in holdings. AWS controls roughly a third of all cloud infrastructure spending and is running at $132 billion annually. They're aggressively deploying AI and LLM solutions to push AWS margins even higher. What's wild is Amazon stock is actually historically inexpensive right now when you look at cash flow multiples - investors can buy at 12x forecasted 2026 cash flow compared to the 30x median from the 2010s.

So what's the actual lesson here about the best way to invest in AI? It's not picking the flashiest AI companies. It's finding dominant platforms that are integrating AI into already-profitable businesses. These three stocks represent about $312 billion in total portfolio value, and the AI angle is almost secondary to the fact that they're oligopolies with pricing power, strong management, and the cash flow to keep buying back stock.

That's how you actually invest in artificial intelligence long-term. Not chasing the newest AI startup, but understanding which established giants have the moat and capital discipline to own the AI infrastructure layer. Buffett's been doing this for 60+ years, and his approach to finding the best way to invest in AI is basically the same as his approach to everything else - find the business with the unbreakable competitive advantage and let time do the work.