TIL about the concept of an "approximate theory".
An approximate theory is one which works, until you look into it too deeply.
It's like flat earth theory. It's plausible (to anyone who never traveled). It can be incredibly convincing. It is "good enough" for most people's needs. It just falls apart the moment you actually start to apply real scientific study to it.
This has a lot of implications in everyday life. For example...
🧵1/6
Commercial software projects eventually become sufficiently complex that normal humans can't understand - or even know - everything that's going on in them. This applies to the programmers working on them AND the users of the software.
At one co. our users had built up an approximate theory of how you had to accomplish a specific task. They all felt it was stupid, but it was never-the-less how you had to do the thing in order to succeed. This knowledge was passed on from user to user.
🧵2/6
The infamous "Story about Magic" was an approximate theory that worked. https://users.cs.utah.edu/~elb/folklore/magic.html The switch was "magic" and if you turned it from the "more magic" to "magic" position the computer crashed. This switch "couldn't" have worked because it only had one wire, but it did work.
You could say that there was an approximate theory that what switch (which couldn't possibly work) was required in order to make the computer function.
🧵4/6
masukomi