I was thinking about Accelerando the other day: in 2026 it's clear that it predicted a lot of things bizarrely well, but that contrasts with Stross' later disbelief that our current world's technology could exist, which seems to have persisted into the current day.