4 ms·
In my own experience, some optimizations seem to come along on their own accord if I repeat a task (constructing and refactoring a code base using a known langu
by osteele 13y ago
In my own experience, some optimizations seem to come along on their own accord if I repeat a task (constructing and refactoring a code base using a known language and platform, keeping a meeting on track, getting the timing and finger angle pressure and muting right while playing a guitar piece), so long as I'm well-rested and attentive during the task, and repeat it enough times.
Other improvements only occur if I consciously decide to try something different (trying a different programming paradigm, changing the cadence or invitation of meetings, using a different voicing or fingering or replacing picked notes by legato or vice versa).
The optimizations that take care of themselves feel more analog. I suspect that they're more susceptible to local optimization. In the case of music (and running, and juggling), I suspect there's some involvement from the cerebellum.
The ones I have to direct feel more quantized; more digital. Optimizing them seems like simulated annealing. Sometimes I learn by knocking myself out of my comfort zone, or copying someone I admire or whose artifacts or performance I admire, or trying out gratuitous variations on approaches to something I already know how to do (I'll write in it Haskell! I'll avoid using my left index finger for a week!) to see what they lead me to.
I don't know if this is the same distinction as the one between progressing within a level versus going to the next level, but that's what I was reminded of when I read this.
The things that I've felt have taken me to the next level are those where I kick myself out of my comfort zone; often in a toy or side project where I've given myself the permission to take huge risks that I don't feel responsible with on a paying gig and/or when other people are depending on me; and often by re-working, in a medium I'm uncomfortable with, an application or problem domain that I've mastered with my old skill set