Almost all the learning you do is learning with a deadline attached. You need the thing for the ticket in front of you, so you read exactly enough to finish, and then you stop. That is efficient and it works, and if it is the only kind of learning you do it will quietly cap you.
Learning under deadline has a shape to it. You skim for the answer. You copy the pattern that unblocks you. You never find out why it works, because the moment it works you have somewhere else to be. You build a fast, wide, shallow layer of knowledge which is genuinely useful and which collapses the first time the situation is slightly different from the example.
So keep a small amount of your learning uncommitted. An hour a week, on something with no ticket behind it and no obvious payoff this quarter.
Read a paper about how databases handle concurrency when you are not currently fighting a lock. Work through a language that thinks differently from yours, not to use it, just to see the shape of another idea. Build a toy version of something you use daily. Read the design docs for a system you will never operate.
The reason this pays is that useful knowledge does not arrive labelled. What you learn with no use yet sits there until, eighteen months later, a problem lands that looks exactly like the odd thing you read about on a quiet Thursday. You will be the person in the room who says, I think this is the same problem as, and half the team will assume you are quick. You were not quick. You were early.
There is a second effect that matters more. Deadline learning teaches you answers. Unforced learning teaches you the shape of a field, which is what lets you ask the right question later and know roughly where to look.
Keep it small so it survives busy weeks. Keep it genuinely interesting to you, because nothing else will get you to the end of it.
The learning with no immediate use is the only kind that is still there when your current job description is not.
– Asael Shinder
Top comments (0)