What Revenue Does the AI Spending Boom Assume?

Scott Galloway says the existential risk in the AI buildout is that the capital spending already committed assumes the market grows to $3 trillion in revenue, while all of big tech is $1.5 trillion in revenue today. That would mean AI triples current big-tech spending, through applications and business services.

He compares the mood to Janet Yellen's line that the risk of overspending was dwarfed by the risk of underspending. The companies, he says, are afraid of missing AI, and only a handful can spend at this scale, so the capital itself works as a moat. He puts Meta's spending on the order of $40 billion and asks what other sector could put three quarters of a trillion dollars into capital expenditure. The firms are also trying to move a lot of that spending onto other investors in private credit, inside separate entities, so a default would not hit their own equity. The railroad, he says, survived and changed the world with a ton of bankruptcies along the way, and a lot of people expect the same pattern here.