Post

I mean I continue to point out that the people who run this website being utterly in the tank for ~*~Agentic Coding~*~ has not in fact led to vast improvements in the overall quality of the service but it *has* led to them becoming 900% more annoying, so idk why I'm supposed to not hate this shit

dame@dame.is

a bluesky user replied to @bsky38.com and said "I don't really want to log into anything powered by Claude AI" dear reader, the Bluesky app is also "powered by Claude AI" 😭

"Oh you don't like vibe-coded single-serve websites built specifically to go viral on bluesky? And yet you post on vibe-coded platform Bluesky, very interesting..."

22 reposts403 likes

I am less annoyed by, eg, the Linux kernel having agents fucking with it (even if I'm not a fan) because Linus Torvalds isn't personally responsible for banning posters I think are interesting because they were mean to their personal friends

This type of "oh, you surely don't *understand it* if you hate it" rhetoric is copied straight from the crypto people a few years ago, who I know for a fact many of you were at the time. I'm perfectly capable of understanding the game mechanics of your stupid sci-fi larp, thanks

Ultimately you can yell "industry standard tools" until you're blue on the face but again, I care about the balance of utility and externalities and that balance is comically in the red. You *are* assholes for using it, and you will *always* be assholes for using it. Get used to it, I guess.

Get used to being hated and scorned. Your already increasingly reviled profession is now hitching its wagon to a technology that amounts to a DDoS attack on the concept of human civilization. You don't get to complain about people seeing your amoral politics for what they are.

They want the Anti-genAI argument to be "it's ontologically evil" because that's an easy strawman. "Technology ought to increase human agency, not reduce it" is much harder for them to argue against, because it leads to the acknowledgement that it reduces agency far more often than it increases it.

Also, they are NOT industry standard tools — they are still experimental tools by any useful metric and they can and do fail spectacularly for no reason that can be usefully diagnosed.

It's also worth pointing out that Linus basically showed understanding himself and put responsibility on the submitters. When asked, he in no uncertain terms told people that if you submit slop commits, the submission rules still apply to you.