On the discourse

It starts with "wouldn't it be cool if." Then it snowballs. There is something beautiful and dangerous about being able to go down the rabbit hole this way. Hundreds of hours spent. Learning about things most would never want to care about. Only the C & V keys know the pain. Looking back now (deeply for the first time) at some of it in totality, the luster is totally gone. From the outside this snowman probably looks more like a desperate grasp for control in an otherwise unworldly situation developing around us.
In many ways it will be effective in providing control. If not at least visibility. For that there is no doubt. Imagine having definitive visibility that over the last three years you crash out in December. Not just feelings but quantifiably: fewer texts to friends, terse messages to the LLMs, sleepless nights, later nights with bad recovery, more PODCASTS than music (gross). All graphed, with significance tests. What do we do with the legibility of this type of data? Hopefully learn from it. Hopefully change it.
There are things you don't want to confront or learn in that process. Even after reflection, things you were way happier not knowing. It’s a self-built pixelated mirror and now you're stuck looking into it.
Why shouldn't you be able to connect email, three to-do lists, calendar, and late night ramblings about how to solve a client problem? Why should software be the friction? But once everything's connected, nothing is separate. Is that worth it? But in the moment of building it all... why NOT bring it all together? Feature drift birthes the panopticon.
So what’s unlocked really? Forward momentum, mostly. The feeling that you can try something by tomorrow instead of someday. More ideas tested in the past year than the previous five. Most should have stayed “someday”. Some became real things that real people use. Whether that's wisdom or waste is still TBD.
All of this at the cost of presence (probably). The constant itch to optimize, connect, measure. That itch can be scratched at the speed of light. Things that couldn't be done before. The army of data scientists in your pocket is real. The dozens of reports to inform a hunch is real. But you scratch one itch and three more appear.
Something shifts in how problems get approached. Not better but different. You start reading about computational thinking. It's hard to NOT feel like you’ve been the thing rewired.
The trip wires are almost universal with non-technical folks. New systems (github). New interfaces (claude code and the dreaded monospace). Terrible UIs (ffs try setting up google oAuth).
Who actually gets to do this? Probably not too many parents or folks working two jobs. Folks with too much free time on their hands, not billing at 110% hours.
I am not worried about technical obsolescence invalidating what's been done. Learning how to be clear in articulating what you want is the skill. We also have a feedback loop that clearly shows you when you don’t at a pace we didn’t before.
More ideas? Yes of course. But when you have spent so much time “co-creating” it makes you more wary. The validation hits different when you know it's cheap. Are the ideas better though.. Knowing about more things (capabilities, reference points) makes ideas better, brings forth opportunities that wouldn’t otherwise be there. And it's easier to validate a segment of those ideas, from hunch to proof, finding benchmarks and research.
It’s hard to shake the idea that depth still comes the old fashioned way. A big goal this year is to read way more fiction, literature, and history. Way LESS industry-related "content" (replacing it with academic or long form work).
A weird sixth sense develops for when the model is confidently wrong. Hallucination spotting..
Maybe all of this is just an elaborate way of having something to do with your hands while things you can't control unfold around you. A grasp for something real as the pace of this otherworld expands faster than our ability to understand it.