I'm a scientist. In the science world, the traditional lab notebook contained a narrative of what you were doing. You're kind of thinking out loud into it.

One measure of a good notebook is if it contains sufficient information that you don't have to repeat work only because you can't figure out what you did. There are other good reasons for repeating things of course.

My spouse is a lab scientist, and I've seen her meticulous notebooks. She was telling me just last week that one of her experiments produced a puzzling result. The next day she said: "I figured it out from my notebook. I skipped a step that was in the procedure."

There was a time when a notebook was also a legal document, and so there was a criterion of whether it would stand up in court as proof that you had invented something. Could a "person skilled in the art" replicate your work based on your notebook? My dad told me that his notebooks were regularly reviewed and witnessed.

The legal issues have changed, since the patent system has switched to the "first to file" rule. My employer got rid of its formal notebook policy when this change came through.

My problem with physical notebooks is that a great deal of my work is computational, and I automate things. In my case, the best form for recording my work is in fact a Jupyter notebook. On the other hand, I come from a family of chemists, and taking electronic notes in a "wet" chemistry lab is often impractical.

I also keep a 'lab notebook', but I must admit that a lot of what I used to document in my notes (setting up software/compiling 3rd party deps) I now document in code (scripts, devops, etc).

I still find lots of value in keeping notes though! And sometimes miss it when I didn't keep notes.