"Please summarize this essentials of this discussion in a way that future agents will understand and put it into AGENTS.md"
...and replying to a sibling; yes, I did add it to `.gitignore` (but that's not a guarantee of it going crazy again), and was super surprised that it truly deleted it rather than "safely" doing `mv ... .trash/*` or something.
The reason to dig into the agent reasoning is that I have to treat myself as if I were the one in error (which as you pointed out, I was!), and determine the cause of it along with prevention.
I have always been an AI quality sceptic. I don't believe quality software is coming out of these models, but I figured it would probably speed up the development of poor quality software. What's surprising is that is seems to not even be doing that. We've had three years of claimed "3x-10x productivity" which means thousands of people have had 9-30 developer years to do something. Where is that output? I haven't seen a single AI developed thing reach Show HN or anywhere else that was worth a damn.
So at this point I have to just assume this shit doesn't work very well for some reason, because no one is outputting anything with it that resembles good, useful software.
As someone working somewhere very much like this, the "everyone" mentioned is actually a few people who are under the mistaken impression that the rest, keeping their head down, are equally interested and on board.
Your post was written almost verbatim by my coworker last week, who has no idea that I and half the team are not doing any of this stuff.
Most things in Remeda, ramda, rxjs, the methods in the Ruby stdlib, etc. would all be great to have. I use at least Remeda in every project when I can.
reply