computation

Computation is not the same as using a computer, and separating the two took me longer than I would like to admit. You can draw in software for ten years and never compute anything.

The difference is what you hand over. Drawing means deciding where the wall goes. Computing means describing the rule that decides where the wall goes, and then living with what that rule does at the eightieth wall, when you are no longer watching.

Which sounds like losing control and is mostly the opposite. A rule can be argued with. You hand it to someone, they find the case you did not think of, and the disagreement is about reasoning instead of taste. Try having that argument about a line.

The catch is the one every optimiser teaches you eventually, and evolutionary search is where I first met it. On lEgoarCh a brick model came back satisfying every constraint it had been given: connected, supported, made of parts that exist. It had also stopped reading as architecture. The machine had not failed. It had done exactly what we asked, and we had described the problem badly.

So computation moves the work upstream, into saying what you actually want, which turns out to be the hardest sentence anyone in this field ever has to write. Is that still drawing?