Stacking Ignorance

Keeping current is hard. It's even harder remembering what you were current on last month. A recent post that flitted by reminded me about Obsidian Vaults, Claude, and the fervor of second brains - claude-obsidian.

There are parts of that concept that make sense, but I found myself falling into the trap of using Claude to curate my knowledge base more often than putting the second brain to work creating output. Arguably this could just be a me problem, I'm not a font of infinite ideas just waiting to be built. It has a similar ring to my tepid view of frontier models, at least in terms of code generation.

Over the last 18 months I've felt an acceleration of a trend that emerged under the miasma of Covid. The rate at which adaptation has been necessary has conditioned me into a loop of Read-Try-Adopt-Supplant where I seem to constantly be foraging for the next what's-it I need to be expert in. This urge to continually morph has spread from development, to playing guitar, to reading, really to every aspect of what I do.

There's a risk I'm falling into a habit of stacking ignorance in the pursuit of currency. For my guitar practice it's much harder to learn to play a song well when I'm spending 5 minutes each on 10 songs. Even practicing daily only certain aspects of my playing get better: finger strength, dexterity, and fluidity; but progress on the individual songs quickly plateaus.

The critical phase of learning that I'm skimming over is focusing on the mistakes. Maybe the chord progression is correct, but the transitions aren't swung the way they should be. Mistakes are the fascinating part of the process, they might be something we want to fix, or they might show us where we can find our meaning in the melody.

A phrase that's stuck with me since the turn of the century is "All things in orbit". It was used as a shorthand for the idea that problems tend to orbit problems. In an organization, if there's a bad leader, you will almost certainly see a solar system of anti-patterns around that leader. If there's a deeply nested hack in a code base, you'll often find circling code levees meant to hold back any fast floods.

For the last few months, the center of gravity around which all of my quandaries have orbited is Value, as in What Value will I see from being current on X? What Value will this spend create for me? What Value does this create for the user?

Value is incredibly nebulous. For me it doesn't have to refine down to a simple equation. The point is to ask and answer the question. The brake I need to apply to my desire to consume all-the-things is the mental lift of defining why it might be of value to me or to whatever project I'm involved in. The how of technology has become impossibly ephemeral; the what is now commodity; the why and its dissection are the practice, the scales, progressions that I need to hone. My goal this quarter is to focus on interrogating value instead of collecting badges, one song, one focus because I want to make it mine.