it helps to be on board with the quasi-positivist idea that the sets of things that can and can’t be done, by any experiment, are not just random collections of engineering details; rather, they’re the scaffolding on which we need to construct our understanding of what’s real.

Scott Aaronson on memcomputing