Hedge fund manager Gavin Baker says investors have been too quick to judge Big Tech's AI spending

Hedge fund manager Gavin Baker said on the 'Invest Like The Best' podcast that investors have been too quick to judge Big Tech's AI infrastructure spending, citing improving cash flow, continued demand for AI, and underpriced compute contracts. Baker noted Meta's free cash flow plunged 91% year over year to $784 million in Q2 2026, but its net cash from operating activities rose 25%, and he argued that high GPU prices indicate strong demand.

Hedge fund manager Gavin Baker, who famously invested early in SpaceX, says investors have been too quick to judge how much Big Tech is spending on AI infrastructure https://www.businessinsider.com/jim-paulsen-wall-street-ai-capex-boom-earnings-stocks-economy-2026-7 . Baker, speaking on the "Invest Like The Best" podcast, said improving cash flow, continued demand for AI, and underpriced compute contracts are all reasons investors should worry less about Big Tech's eye-popping spending. "You're going to see a lot of acceleration that's going to answer these ROI questions," Baker said, referring to what he says is a recent increase in the amount of cash these AI businesses are collecting. "You've started to see that this quarter." In July, investors rushed to sell memory and chip stocks. They were largely worried that companies like Meta were spending too much https://www.businessinsider.com/meta-q2-2026-earnings-analyst-ai-strategy-throwing-spaghetti-2026-7 on AI infrastructure without a clear path to generating returns on those investments. Meta's free cash flow https://www.businessinsider.com/meta-q2-2026-earnings-free-cash-flow-plunges-ai-investment-2026-7 plunged 91% year over year, to $784 million in the second quarter of 2026. The company, however, posted a 25% year-over-year rise in net cash from its operating activities. "That's a material acceleration at this scale," Baker said. Baker also said that persistently high GPU prices show that demand for AI computing remains strong. "I don't think anyone in '24 or '25 thought that the prices of old GPUs would still be going vertical in 2026," he said. Many cloud providers, Baker said, have yet to fully monetize the AI infrastructure they've already built because much of their compute capacity is tied to older, lower-priced contracts they signed when GPUs were cheaper. "Essentially, you have the contracted base of installed compute trading at a massive discount to the current spot market," he said.