yeah that thought process is an application of jevon's paradox except the thing is actually artificially cheap and trying to take advantage might just blow up if it doesn't become actually cheap

yeah writing tests can help it get there a bit quicker but this is all true. i think the key to this model for a lot of people is just "well what if we didn't care about code quality or readability. it's like how we don't care about assembly anymore mostly. just throw it away and start over"

that said, there's another model of thinking where if the cost of generating code drops to near 0, then you can do iterative waterfall. design, generate, update design based on feedback, regenerate. generating codebases where functionally no one understands them doesn't feel great to me though

this is actually part of a strategy. i was 'off' 'social media' but actually i just replaced my old addiction with scrolling reddit, which is somehow worse for my brain. new strategy is that i can have a little, as a treat, and it shares a time budget with reddit

i've been watching it grow for a month! feeling lucky it got most of the way there before the snow and was able to finish ripening inside

apparently even for house cats vet manuals will say stuff like "do not get in a fight with a house cat, you will lose. they do not fear death and do not care if they hurt you"

"ah yes so ferocious" i say patronizingly, knowing full well i would lose in a fight

fingers crossed forever