Ask HN: How do you manage self-study?

Practicing a musical instrument, falling while snowboarding, getting up early to be on the mountain for the sunrise.

The original poster is asking for ideas to help build enough discipline to power through the tough spots of a learning curve to get to where the proficiency pays off in enjoyment or other benefits.

We have so many easy diversions, that it's easy to train one's self to not do hard things.

I can study/practice/work for a future benefit, or I can play a video game (or browse hn) for fun now!

Training yourself to do hard things is harder. Maintaining a vision of the goal, getting satisfaction from expending effort, having empathy for your future self, denying yourself of immediate distractions/pleasures can each play a role.

One issue with learning things is that as you learn, you often discover new things to learn. It's critical to know how to file these: (1) they're true dependencies, (2) they're interesting follow-ups for later, (3) they're not relevant in the near or medium term.

File things aggressively. Try to identify as many things as you can to not read. For this you need two things: (a) a clear understanding of where you're going, based on a plan working backward from some achievable thing you want to learn and (b) safety knowing that when you file something away, you won't lose it. It's interesting, and exciting, so you should return.

Do this well and (a) and (b) reinforce one another. You can look through your "set aside" list to build out a strategy for your next achievable task and your achievable tasks can give structure to the things you've set aside.

File things aggressively. This often feels bad. You feel like you "ought" to know something or you get carried away following a thread. That's not bad! Do it sometimes, learning playfully is fun! On the other hand, you'll be more satisfied 3 months from know if you learn aggressively and directedly. You'll have worked through a greater amount of material at greater depth.

Finally, it's often hard to get started. Don't overthink it. Just set a small achieveable project, maybe give yourself a tight deadline. Once you're through that project you will, without a doubt, have a much clearer idea of what comes next.

When I wanted to be a better sysadmin, I forced myself to use desktop Linux back in 1997 when it wasn't nearly as easy as it is now. I had to learn how to compile and configure kernels, how to manage drivers and displays, how to write scrips, a bunch of hardware internals so I could configure them correctly, etc.

But I was driven by the overarching goal of having a usable desktop machine.

You can do the same here. Pick a larger project that accomplishes something you really want to do, and then learn what you need to learn.

You said you have an interest in ML -- build an image classifier off of a camera feed at your front door. Make it identify the mail carrier and your neighbors' cars. Write it in a language you don't know but want to learn. And so on.

Your progress will be tracked by how satisfied you are with the project and if it meets your needs.

In general however, my advice is to relax. There is more to life than knowing little bits of many things. Try to learn something outside of computers. I find building things with my hands that aren't on a screen a lot more gratifying generally. Also, those kinds of skills come in a lot more handy in a zombie apocalypse. So there is that.

1) Every weekday 30min uninterrupted study at a specific time, preferably in the morning. Don't miss, rain or shine. Take weekends off, though.

2) Work in a strict linear fashion: read a chapter, then solve all of that chapter's exercises in order without skipping any, and only then allow yourself to peek at the next chapter. Don't "take a first pass through the book", none of that. To remember where you are, use a bookmark.

3) If you're having trouble with some exercise, you can look up the answer key for that specific exercise, but only after you've spent 5 minutes of effort on that specific exercise without making progress. After that, the same rule applies to the next exercise.

4) You've got to get to the end of the book. Matter of pride.

For me this approach has worked well for established topics in math, physics, econ. Haven't tried it for CS or ML.