Greenspan called it "irrational exuberance" in Dec '96 and the Nasdaq still nearly 4x'd before topping, so being right about the bubble but early is just being wrong with extra steps. Amazon dumped ~90% in the crash and still became one of the best trades of the century, while META and ORCL print tens of billions in cash flow, they're not pets.com. Doom posts always sound smarter, but the money goes to whoever can sit through the ugly headlines, not to whoever predicts them.
The thread's argument is about financing costs: the claim that money-losing AI builders now face double-digit borrowing costs while profitable ones keep compounding. The top replies push back that being early on a bubble call is still being wrong.