Bank a Win Then Figure Out How You Did It
by Mike LevinWednesday, October 26, 2022
I do a lot of writing in SimpleNote in the bathtub and stuff that normally doesn’t reach here. These are mostly unfinished pieces, but none-the-less would still appear here if I just did the original writing in my journal. But I’m trying to keep SimpleNote clear of everything except what’s new and top-of-mind, so occasionally I’m going to “dump” some of that content here. Here’s one.
The world has greatly turned your way. What it values is different than when you were young. Your introversion and nerdiness are currently well-paired with the tools that are out. It still feels like it should have been Amiga, but Unix via Linux on fairly cool hardware will do. Time is running out and nothing evolved is all that well optimized. It has more of a get ‘er done attitude. Reminds me of the 80/20-rule. Bank a win.
Bank a win then figure out how you did it. Make it reproducible, and preferably portable between hardware platforms. Don’t allow your know-how to be tied to a time, place or particular hardware platform. You can’t get away from being tied to a software platform because you need to learn something to operate the hardware. Application program interfaces, or APIs, are simplified languages for talking to hardware. You learn APIs. APIs should be very portable.
Desktop and mobile point and click software is quite mobile if licensing issues like Apples aren’t stopping you. But graphics-based software is uniquely tied to hardware because of the need for a high-resolution pointing device. This very much ties software that focuses on the human user interface to the hardware of the day and sets the stage for your skills going obsolete has that hardware platform goes out of fashion and out of production.
So we switch tools. We re-learn the new human interface software for new hardware and a new platform. We pay a new cloud overlords for old things done in new ways. Those good at this process, switching tools and never falling in love with specific medium and craft, excel. They can flit from language to language and API to API as if butterflies to flowers. Oh how I envy them. I am not one of them. I like to settle in on tools for awhile, getting to know their ins and outs, nuances and forever improve.
Some people are naturally multi-lingual and can talk to different hardware easily. I’m jealous of them. Others take shortcuts seizing on single ways of doing a thing and shut their minds, sometimes violently, against the new. This can work for them and they can protect turf and “widen the moat” in Warren Buffett terms and I’m a bit jealous of them too. But here in the middle is actually not a bad place to be if the tools are cool and you can fall in love with your craft.
I know I love using language to express but am not naturally multi-lingual nor all that quick on the uptake. New abstractions come slowly to me and telling one from another is difficult, such as between Algebra and Calculus. Yet I’m pinning my career and livelihood exactly on such a selection of tools and language that will take me forever to master. That such Goldilocks-tools even today exist today balanced between the top-technical and too-simple is a miracle.
Even if one day you can speak with an AI that is capable of doing almost anything and everything better than you including coding, it is still good to learn coding. It is for the same reason that it is still good to learn to drive. Not doing so is tantamount to giving up.
And I choose this path. I choose it because it enables me to hold discourse in the precise language of technology.
Many will skip learning how to program computers or even to drive, figuring of you just wait long enough, you’ll always have someone to do it for you and you’ll always have a ride because AI. If you look carefully at that situation, they’re waiting for either servants or surrogate parents. They’re trying to remain in a child-like state. A life of luxury, free time and fattening up like a Thanksgiving turkey.