What if Humanity forgot how to make CPUs?

Its an interesting thought experiment, but not a very coherent one, as the author makes an unfortunate equivalence between ultra-complex digital designs and all silicon ICs. Losing the latter is basically impossible, as it's possible to make ICs with <100 transistors at home, which means you can always build simple digital designs with discrete circuits. CPUs in general are neither classified like FOGBANK, products of particular weird metallurgy like the F1 rocket, nor catastrophically unprofitable to produce like CRTs; someone would just start making them again at whatever the available scale was, including discrete transistor scale, and then shrink from there.

I think it's an interesting thought experiment that invites each participant to imagine how long they could keep their own technology running.

As someone who runs old m68k and VAX machines 24/7, I have a lot of experience about what goes wrong and what to do about it, so it's interesting to see other perspectives and guesses. I don't think Laurie was wrong, but I do think her timeline was a bit shorter than it would be in reality. After all, even people who aren't preserving '80s and '90s machines are running old iPhones and computers on which people will never run Windows 11.

I think one point worth discussing is how much waste there is, both in the landfilling of perfectly good hardware and in the waste of computing resources. If anyone needs proof of how much the latter is unnecessary, look at how much can be done on the modern Internet with an Amiga.