Financial experts are advising investors to pivot from traditional AI stocks toward companies building the physical infrastructure that powers artificial intelligence [1].
This shift matters because the rapid expansion of AI workloads is creating an urgent demand for the hardware and energy systems required to run complex models. These infrastructure providers may offer more stable, recurring revenue streams than the volatile software and chip sectors [1, 2].
In a recent discussion, Kenny Polcari, Jared Blikre of Yahoo Finance, and Michael Monaghan of Founder ETFs said that the next phase of the AI boom lies in the "toll-takers" of the industry [1]. This includes firms specializing in power generation, data center construction, cooling systems, and optical components [1, 2].
While chip makers like Nvidia have dominated early investment cycles, some analysts suggest the market is shifting. Some reports indicate that while Nvidia remains a dominant player, it is facing increased competition [3]. Other market perspectives suggest that buying infrastructure stocks, such as Vertiv and EME, is a more strategic move than focusing on the "Mag 7" or AI chip makers [4].
Beyond the data center, the broader tech ecosystem continues to see massive valuations. For example, SpaceX has been cited with a valuation of $1.77 trillion [5] and an IPO share price of $135 [5]. The company reported revenue of $18.7 billion in 2026 [5].
However, the core argument from the Yahoo Finance panel remains focused on the physical layer of AI [1]. They said that without the underlying power and cooling infrastructure, the software and chips cannot function. This creates a bottleneck that benefits the companies providing the essential physical components [1, 2].
“Investors are being urged to prioritize AI-infrastructure firms over traditional AI chip or software stocks.”
The transition from AI software to AI infrastructure represents a move from speculative growth toward the 'picks and shovels' of the digital age. As AI models grow in size, the limiting factor is no longer just algorithmic efficiency, but the physical capacity of the power grid and the thermal limits of data centers. Investors are now hedging against the potential saturation of chip markets by betting on the indispensable utility providers that make the entire ecosystem possible.
