Big Tech’s cash burn might actually be a moat build

Not bleeding cash — redeploying it

Investors have been side-eyeing Big Tech’s AI bill like it’s a college kid’s first credit card. But Chamath Palihapitiya argues that lower free cash flow at Alphabet, Meta, and Microsoft is mostly a math problem, not a business problem. Operating cash flow is still strong; the difference is they’re shoveling more of it into data centers, chips, networking gear, and power systems.

The Amazon playbook, but with GPUs

His comparison point is Amazon, which spent years stuffing money into logistics and AWS before those investments became a monster moat. In Chamath’s view, the hyperscalers could be doing the same thing now with AI: taking a short-term free cash flow hit to build infrastructure that’s very hard for rivals to copy later.

Who gets paid if he’s right?

This is where the story stops being just a philosophical argument and starts smelling like a supply-chain trade. If Big Tech keeps spending hundreds of billions on AI infrastructure, the money doesn’t just sit with the megacaps. It ripples outward to Nvidia, Broadcom, TSMC, and the network/power/data-center ecosystem that sells the shovels for the AI gold rush.

Big picture

So no, the headline here isn’t “cash is disappearing.” It’s “cash is getting converted into a moat.” Whether that turns into a brilliant long-term setup or an overbuilt science project is still the trillion-dollar question.