in case you need a taste of how fucked the tech industry is right now, I'm being required to use AI at work. if I talk about how it fucks up or overcomplicates basic asks, it's because I "don't know how to use it" which indicates a "lack of growth mindset", and thus poor performance. I've been told this directly to my face, starting immediately.
so not only must I use Claude, I have to cover for Claude's mistakes, and then go the extra mile to pass off my own work as Claude's.
anyway, I need to pay rent, so I guess I'm a slopper now.
at this point, I want to move to a farm and shovel pig shit for the rest of my life.
@AmyZenunim I'm so sorry. That's such bullshit.
This kind of shit is why I'm trying to get the pre-reqs I need to get into a nursing program. I'm just done with professional computer touching at this point.
I hope you're eventually able to find a job that doesn't force this kind of horse shit on you. Or if you're actually serious about the farm thing, I do know some pretty rad resources for aspiring farmers I could send your way?
Note: LLMs are filling the healthcare field with dangerously misleading nonsense that kills people. 💢
@AmyZenunim I wonder if this will get tech workers to develop greater class consciousness or no.
BTW, is there a list somewhere of companies using AI (internally or not) so I can avoid them? I really don't want to be taking on the risks associated with poorly-engineered/unengineered technology.
For your sake, I hope you keep answering honestly (and resisting), or find something else. It's not your growth they care about.
Oppression is being told how to think.
> I have my doubts LLMs are actually speeding things up.
The expectation is to see more patches co-authored by Claude or similar, which completes stories and deliver something, even if it's incomplete, buggy or plainly wrong. It doesn't matter how many new specs are being added or how confident are you in the result; what it matters is getting things out of the door - push it, crush it.
That's kind of the never-ending cycle at least for software developer context.
@AmyZenunim
Same.
They're paying for my time, if they want to waste it I guess that's their business
@AmyZenunim
Same. I'm looking for something better, but it's a slog through endless other slop shops, competing with other good devs for the few organic dev jobs still out there.
Meanwhile I've gotta eat to live and gotta work to eat, so slop it is.
@AmyZenunim this sound suspiciously familiar. It's either terribly common or we share an employer 🥲
Anyway, moving to farming is exactly what I'm doing now, shoveling shit gives more satisfaction to me than forced bot interlocution.
@AmyZenunim I do not understand this. Don’t know about the rest of you, but I’m perfectly capable of getting things severely wrong by my own accord.
Why do I suddenly (after 20 years on the job) need a subscription to environmentally horrific auto complete trained on theft?
They really are just trying to kill off the software dev as a profession, huh. Even if the bots can't replace you, they're trying to cripple your skills as well. Wishing you the best in making it out of this nightmare with all your capabilities intact.
and THEN, when Claude injects some shit into your product that breaks it with catastrophic results, you'll get dumped on for not reviewing Claude's work sufficiently well, to which you respond "If I have to review its work, how does that save time?" to why they'll tell you to shut up.
I feel ya. Im right there with ya.
i said review takes time. they said “skill issue” 
@AmyZenunim ...and shortly after I made that reply, one of my coworkers PRed a giant mass of vibecode. Thankfully it's almost entirely DTOs in a brand new package, so it couldn't mangle things that badly, but it's clear from the comments and (lack of) naming "sense" that he just let the "AI" run wild without really understanding anything it was doing. Certainly doesn't help that his programming fundamentals were somewhat lacking to begin with, though. (I noticed him starting to lean on it a while back and warned him it was a bad idea. Looks like he hasn't listened...)
(I can also Just Tell that it wasn't him who wrote most of it. It's not written in his usual style at all. There was no real thought process, only pulling the handle on the fruit machine and committing the result if it looked like it got things closer to working. And yes, I can tell that just from the code. It's soulless, the thought patterns I can usually see just aren't there)
@AmyZenunim have you checked for loopholes?...
AI is not explicit Neural Networks or Large Language Models for that matter, it's just a branch of all the intelligent machines we have.
Just bring a thermostat to work and call it a day. Thats feedback loops and those are AI.
@AmyZenunim right there with you. CTO and COO have been pushing AI at my work place. The result? Merge requests with a dozen useless commits, and everyone barely has a clue what's what. But it's okay, we've got Claude reviewing the code that Claude produces, not sure about the result? Just keep running it over and over rather than actually looking at anything, wheeeeee...
Yeah, it sucks. There's no regard for quality.
@AmyZenunim I have this at work too :/
I'm hoping it's some kind of mass delusion that passes when the bubble pops and no one can afford it any more!
Is "tell the company AI to do something token-using in the background, get on with doing normal accurate work yourself" an option?
(I'd say "something that provides actual value to somebody else", but examples are hard to come up with. Maybe have it write fanfiction? At least there the errors are funny.)
There's more than enough Generative AI slop "fan fiction" driving the public social media sites these days.
@ErinPtah
@AmyZenunim
I think "tell the company AI to do something token-using in the background" would be called "agentic", which would probably count as even better than using it directly
One could even set it working on the actual code + ticket, in case there are spot checks
@AmyZenunim I just heard almost the same story, just this morning, from a friend in Poland. He mentioned how his programming skills were atrophying because he's just an assistant for the AI.
Which makes me wonder: If you're coding all the time, how will you adequately review the AI code for errors?