T. Rowe Price Looks Beyond Chips for AI's Next Winners

The next AI beneficiaries will come from healthcare, robotics, financials, and industrials, if T. Rowe Price portfolio managers Shaun Currie and Jon Friar are right. On the firm's 'The Angle' podcast, they argued that these sleepers will emerge once the current data-center buildout matures; Currie calls this period a 'data center cycle' rather than an 'AI cycle,' comparing it to the late-1990s internet buildout, a five-to-seven-year hardware phase that gave way to two decades of e-commerce, mobile, and cloud booms. The implication for ETF investors is that the easiest money in AI may already have been made, and the argument redirects attention from the crowded chip-and-cloud hardware trade to the applications layer.

The analogy is more than historical trivia: Currie's point is that AI adoption does not reward size the way earlier technology waves did. A $20 million annual AI budget already places a company in the top 20% of corporate spenders, a bar many smaller firms can clear and one that should drop as AI costs fall. Robotics fits the same thesis—'programming a robot today costs more than building a robot,' so cheaper AI-driven programming could let robots handle multiple tasks, lowering costs and triggering what Currie called a self-reinforcing flywheel. Healthcare and biology data providers, he said, carry less downside than hardware stocks that have already had a massive run.

For T. Rowe Price, the talk doubles as a product pitch: the actively managed T. Rowe Price Technology ETF, run by Dom Rizzo, is the straightforward route into the chip-and-cloud buildout. The sleeper narrative maps more precisely onto the T. Rowe Price Small-Mid Cap ETF (TMSL), managed by Jodi Love; benchmarked to the Russell 2500, the fund already carries a 22% weight in industrials and business services as of June 30 and can range beyond the index on the strength of the firm's proprietary research, suggesting the fund is already placed for the shift its managers describe. Friar's screen for sleepers—technical depth reaching into the executive suite, network effects that lock in users, and physical assets like sensors or logistics networks—is a description of small and mid-cap industrials and healthcare companies.

The term 'sleeper' is doing a lot of work: it implies the market is underpricing these companies, the exact claim an active manager needs to make. The argument extends the case for active sector ETFs that T. Rowe Price's health-care fund has been making from the other side of the market, and it lands at a moment when investors are paying up for active equity strategies. Whether T. Rowe Price can actually find these sleepers is the open question, but the data-center-cycle framing is a specific, falsifiable call: if the buildout is as early as Currie suggests, the next phase of AI winners will be messier, more fragmented, and more forgiving to stock-pickers than the first.