One thing that author forgot: automation not only helps you with ergonomics and changing your behaviours - most important thing is that it prevents common mistakes in repetitive work. Some may argue that it is part of ergonomics, but I think it is mostly separate thing that is often forgotten.
Yes, this omission is pretty ridiculous. The time saved is calculated in an absence of documenting it in a fool proof way. If I document it in a fool proof way with 17 steps of commands to execute, I can also automate it for a fraction of added cost and then I won’t any mistakes any more. And I can reverse engineer the automation if I documented it…
I was bracing to disagree, worried that the author would say something like “automation is a fun and cool buzzword so that makes the time sink worth it!”
I’ve unfortunately encountered those people who will insist on automating some internal process that 1) happens rarely, 2) is quick, and 3) changes a lot. Pointing out that the time savings (across all ~3 users) is not worth it was not effective somehow.