Gavin Baker, the hedge fund manager known for his early bet on SpaceX, believes Wall Street has been far too impatient regarding the astronomical sums Big Tech is pouring into artificial intelligence. Speaking on the Invest Like The Best podcast, Baker argued that critics focusing solely on massive capital expenditures are missing the larger picture, suggesting that it is simply too early to determine whether these investments will pay off.
The skepticism reached a fever pitch in July when traders dumped memory and chip stocks, fearing that giants like Meta were spending billions on infrastructure without a viable way to turn a profit. While Meta saw its free cash flow plummet by ninety one percent year over year in the second quarter of 2026, Baker pointed out that net cash from operating activities actually rose twenty five percent. To him, this represents a significant acceleration in revenue generation that should ease investor anxiety over return on investment.
Beyond the balance sheets, Baker noted that the market for hardware tells a story of relentless demand. He observed that prices for even older graphics processing units continue to climb vertically well into 2026, a trend he claims few analysts predicted. This suggests that the appetite for AI computing power remains hungry despite the narrative of an impending bubble.
Finally, Baker explained that many cloud providers haven’t yet realized the full financial potential of their existing infrastructure due to timing. Much of their current compute capacity is locked into older contracts signed when GPUs were significantly cheaper than they are today. According to Baker, this means there is a hidden reservoir of value waiting to be unlocked as those discounted contracts expire and are replaced by current market rates.
