THE AI SELL-OFF COULD BE AN OPPORTUNITY Wall Street is starting to treat AI investment as a problem. I think that's exactly where it's getting the story wrong. Over the past week, we've watched some of the world's best businesses deliver exceptional results: - $MSFT reported $90B in revenue, beat EPS by nearly 12%, Azure grew 43%, Commercial RPO surged to $678B and Azure crossed $100B in annual revenue for the first time - $GOOG posted another blockbuster quarter, with Cloud revenue up 82% and cloud operating income more than tripling year-over-year - $META grew revenue 28%, reached 3.6B daily active users, and continued to grow advertising pricing, yet the stock sold off after raising 2026 AI infrastructure spending Despite those numbers, investors focused on one thing, CapEx. The narrative has quickly shifted from "AI is the future" to "They're spending too much" but ask yourself...Why are the largest technology companies in the world all increasing investment at the same time? Because demand hasn't slowed: - Microsoft admitted Azure growth is being constrained by available compute capacity, not customer demand - Six-year-old A100 GPUs are still becoming more expensive to rent and the four largest hyperscalers have accumulated nearly $2 trillion of cloud backlog waiting to be delivered That isn't what an oversupplied market looks like. It's what a supply-constrained market looks like. Today's AI spending isn't just another expense on the income statement. It's the infrastructure being built to serve demand that already exists. Over the next 12–24 months, much of that backlog is expected to convert into revenue. As existing contracts renew at higher pricing and infrastructure begins operating at scale, revenue should continue growing while CapEx growth is expected to moderate over time. That's when today's investment starts showing up as tomorrow's free cash flow.