In retrospect, I feel like I lucked into an industry where the work is extraordinarily flexible and disproportionately lucrative.

So once I reached a mature point in my career, I emphasized taking control of my own time so that I could blend other things into my life as I wanted to.

Industry stuff pays the bills and I can make decisions about how big I want those bills to be and how much I want to work to support it. The rest of the time can be spent doing almost anything.

I've moved around within software — from developing a commercial app in the finance industry, to developing simulations of air traffic control for the FAA, to developing distributed a command and control plan view display for the Army, to building datamining a custom supercomputer-based hadoop sized VLDB for social activity for the government, to supporting HPC users in academia, to modeling disease in a pharmaceutical company, to analyzing 3D biomedical images to quantlfy drug effects, increasingly using ML and deep learning.

In all, that spans 32 years. I've found the trick to avoiding boredom is to not stand still in the same problem space for too long.

How's Scrum/Agile/micromanagement in the fields you've mentioned? I've heard that even some research units at universities are beginning to use Agile now...

When developing prototype apps, you tend toward agile practices naturally. Status/update meetings are regular and quick, if not daily. The chief difference may be the absence of Agile's task cards, which requires too much forward decomposition of the design into bite-sized units, IMHO. R&D projects know the starting point and the objective, but only in terms of utility, not the means to that end. You're managing your team's passage through the problem space, not the solution space. That requires a lot of interaction with the client, since you need to learn from them if what you can deliver will serve their need. Thus you need to revise the task cards a lot, to the extent they just aren't that helpful in gauging your progress. It also helps that the team size in R&D is generally tiny, like one or two.

Ski instructor was one I heard of. This was from someone who was already very sporty, so learning to ski then became an instructor in a short amount of time.

I know a young Bay Area engineer who left a large tech company recently to teach Brazilian Jiu Jitsu full-time. He felt that BJJ was the activity that was the most age-senstive to accomplish, not SW.

Degree in Aerospace Engineering. After 6 years of job jumping became a fire fighter / paramedic. Still a FF/PM full time but got a master's in computer science and now work part-time remote as a programmer.

Film making is one I heard of and also aspiring to do.