hayward@click:~/~/rants/2026-10-29-the-asking-used-to-be-public

what these numbers are

Every token my AI tooling has sent and received since 12 Aug 2026, kept in one running total that never resets.

is what goes in, is what comes back out. The in figure runs far larger because most of it is context I re-send with every request rather than new text.

Both tick up live while anything is running, and stop when nothing is being spent.

Counted from the usage my AI services report, and not all of them report tokens, so treat this as a minimum rather than a grand total.

The asking used to be public

29 October 2026

I did not learn this trade from a manual. I learned it from other people's mess, written down in the open. Somebody had the fault before me, lost an evening to it, and left the argument on a page I could find. The wrong answers were still there beside the right one, and the reason the wrong ones were wrong taught me more than the fix did.

Asking was a public act. You put the question somewhere it stayed, people came at it from different directions and disagreed, and the thread settled or it did not. If it never settled you learned which parts nobody knew either. Years later you found your own old question, and the answer a stranger gave you when you were greener, still there and still useful to people you will never meet.

I do not ask like that now. I ask in a box, on my own machine, in a window that closes. The answer comes back in seconds and it is usually better than the thread would have been. It is also the last time that answer exists. Nothing survives the window: nobody reads it afterwards, nothing gets argued over, and being wrong in front of people who would correct you costs nothing.

I have noticed the trail changing direction. I used to search for a problem and find somebody's evening. Now I search and find a summary of everybody's evenings, written by something that never had one. The summary is tidier. It has no argument in it, and the argument was the part I learned from.

What I do about it is small and I am not sure it is worth much. I still write up what I find out. When I chase a fault to its cause on my own box, I write the sequence down, in order, badly, in my own notes, somewhere I will trip over it again. When a repair turns out to rest on something worth knowing, it goes in the file I keep for the machines rather than staying in my head.

Writing it up is the last of the old habit I have kept. It no longer puts anything into the place where the learning happens, because that place is a box that shuts. It puts a copy where nobody argues with it, and an argument is what it would take to find out whether it is true.

I keep thinking about the person who meets this same fault after me and goes looking. The answer will be there, fluent and confident and unargued. Nobody will have left an evening behind it.

All rants · The part of it that lived in my hands