This can be an incredibly powerful tool. It does so many things - it forces you to write down what you're doing. If you actually do that, it's just a great opportunity to think "Hey hang on, is this thing I'm doing actually solving the problem I have" and "Actually is this the most useful thing I could be doing" in an explicit way. Another thing that the author doesn't really mention but actually is really useful for some of us is to go back and read what you've done to (a) not feel like you've been wasting your time doing nothing, and (b) to realise where you were going wrong on this line of reasoning earlier.

Personally I think this is a really useful time to be revisiting the concept because this is how a lot AI tools work. They're language models so the way they get to complexity is through writing out plans step by step and running commands and then interpreting them, you can read your session with an AI agent like claude code as their engineering notebook.