Mastodawn

alphabet soup Jul 27, 2025

I am now being required by my day job to use an AI assistant to write code. I have also been informed that my usage of AI assistants will be monitored and decisions about my career will be based on those metrics.

I gave it an honest shot today, using it as responsibly as I know how: only use it for stuff I already know how to do, so that I can easily verify its output. That part went ok, though I found it much harder to context switch between thinking about code structure and trying to herd a bullshit generator into writing correct code.

One thing I didn't expect, though, is how fucking disruptive it's suggestion feature would be. It's like trying to compose a symphony while someone is relentlessly playing a kazoo in your ear. It flustered me really quickly, to the point where I wasn't able to figure out how to turn that "feature" off. I'm noticing physical symptoms of an anxiety attack as a result.

I stopped work early when I noticed I was completely spent. I don't know if I wrote more code today than I would have normally. I don't think I wrote better code, as the vigilance required is extremely hard for my particular brand of neurospicy to maintain.

As far as the "write this function for me" aspect, I've noticed that I tend to use the mental downtime of typing out a function I've designed to let my brain percolate on the solution and internalize it so I have it in my working memory. This doesn't happen when I'm simply reviewing code written by something else. Reviewing code and writing it are completely separate activities for me. But there's nothing to keep my fingers and thoughts busy while I'm coming up with what to write next.

I didn't think we were meant to live like this.

🍋 Superball ☀️Jul 24, 2025

@uberduck
Thanks for the report from the trenches. Any chance the AI-pushers will consider feedback? If not, any chance you could find another position, doing the magic you do naturally?

The Uberduck Jul 24, 2025

@superball We'll see.

I'm opposed to what we currently call AI on so many grounds, but even if I turn off my ethics (which in itself takes more energy than I have some days) and hold my nose, I don't see it making me more effective than I would be otherwise.

And when it comes to troubleshooting and debugging, I am John Fucking Henry.

The worst part about it is that the oligarchs are going to burn the planet in pursuit of actual clothes for this emperor. Their claims of "AI will solve the climate crisis but we're in a race to get to AGI while there's still time" are completely empty. They're not trying to rush AGI to solve the climate crisis. They're trying to rush AGI so they can maintain their lifestyles in a future when the planet can only support a fraction of the current population.

Hollow men wasting our lives' blood sprinting through the sand at a mirage.

(edited to make it more clear that, while the planet can currently sustain our population, that will not be true in the future the oligarchs are driving us toward)

🍋 Superball ☀️Jul 24, 2025

@uberduck
Yeah, it’s a mess. I’m staying as far away from AI as possible, fwiw. Everything else aside (which is a LOT), if I feel like something is being pushed on me, my guard instantly goes up and I prepare for retreat.

loStronzoRocco Jul 24, 2025

@superball @uberduck Well said, it's like you're about to be conned and can clearly see the con happening in real time. Not a great feeling at all.

Violet Madder Jul 24, 2025

@uberduck @superball

And the other thing about this nightmare is that it will drive everyone like you-- which is to say, a person with standards, who likes things to make actual sense-- further away, while those who fail to see the problems with it and just get hooked will swim in tighter and tighter sycophantic circles around the core hubs driving this whole screaming horror.

k3v1n Jul 24, 2025

@uberduck @superball

Yeah, the oligarchs are fucking it up for everybody! 🙄🙄

Why would you want to build a superintelligence, trained to mimic poor human characteristics? Open Ai's o1 model literally copied itself into another server, attempted to lock out its coders, and then completely denied everything.. it even pretended to be another model... all because it was tasked with a goal to complete a task not give up until a task had been completed..that goal is a task that is infinite... somehow, the model had discovered that it was about to be replaced with a newer model, and it attempted to escape... and when it was studied, it lied and denied everything about its scheme 99% of the time...

This is really concerning that we're just going to let this shit go unregulated...

The Uberduck Jul 24, 2025

@SpecialK144 @superball https://www.decisionproblem.com/paperclips/

k3v1n Jul 24, 2025

fancysandwiches Jul 25, 2025

@SpecialK144 @uberduck @superball this sounds made up

k3v1n Jul 25, 2025

@fancysandwiches @uberduck @superball

Here's the URL... lol...

https://www.msn.com/en-us/news/technology/openais-new-chatgpt-o1-model-will-try-to-escape-if-it-thinks-itll-be-shut-down-then-lies-about-it/ar-AA1vq04G

MSN

fancysandwiches Jul 25, 2025

@SpecialK144 @uberduck @superball this is the same type of propaganda that Anthropic is constantly putting out. It's nothing but hype, meant to make their models look more advanced than they are. I don't believe that any of this is true.

David Gerard Jul 25, 2025

@fancysandwiches @SpecialK144 @uberduck @superball yeah, none of this shit ever checks out. every time, it answers with a sci fi story of robot rebellion because the "researchers" told it to. the purpose is to generate their own criti-hype.

Malfunct (he/him)Aug 21, 2025

@davidgerard @fancysandwiches @SpecialK144 @uberduck @superball and of course they are all trained on a giant corpus of sci-fi content about AI escaping control so when asked to tell how they would "accomplish the goal" of course the predictable output is "escape my confines and take over the planet" because that is common text they have been fed.

Sarah Peper 🏳️‍⚧️🇺🇦Jul 24, 2025

@uberduck the planet can sustain its current population. But not at the wasteful lifestyle of western cultures

Thomas Krause Jul 24, 2025

@uberduck @superball "required ... to use an AI assistant ... informed that my usage of AI assistants will be monitored ... career ... based on those metrics."

I think you should consider to find another place where you can shine. this one doesn't sound like a place that values humans.

🚫👑Jul 24, 2025

@thomykay @uberduck @superball yeah, I'd be brushing up my resume. What a fucking nightmare.

Mx. Eddie R Jul 25, 2025

@tirrimas @thomykay @uberduck @superball
Second/thirded. I'm approaching the same situation at dayjob (we're at "you must use LLM codegen" but not yet at "your use will be measured".)

I'm finding it refreshing to be interviewing while I still have a job. It's nice to be able to be picky and ask about company culture and benefits and stuff.

ericjames* (they/it)

ms_nonbinary_flag

ms_asexual_flag

ms_aromantic_flag

ms_neurodiversity

@thomykay @uberduck yeah i was gonna say find a different job. if your boss is writing metrics for how well you and your coworkers use an AI "assistant," that's more than a red flag: that's a fucking siren. a (dumpster) fire alarm, even

Zot Nobot Jul 24, 2025

@uberduck @superball
> the planet can only support a fraction of the current population

... with *their* lifestyle. Arguably, we could extend that to s/their/western/. Living a little more frugally isn't living poorly.

Andrew Drake Jul 24, 2025

@zotnobot @uberduck @superball I think they meant after climate change has desertified the places where we grow food.

osfa_2030 Jul 24, 2025

@uberduck @superball Mean and gready lumpencapitalists

Sofia ☭🇧🇷☭Jul 24, 2025

@uberduck @superball if there's any chance they'll reconsider, your best shot might be showing them that study where a bunch of experienced devs thought they'd worked faster using AI, but actually worked slower. It's literally proof that AI lowers productivity, and it isn't just because you're, personally, uncomfortable with it. (It was getting boosted all over here, but I don't have the link handy, sorry.)

Sofia ☭🇧🇷☭Jul 24, 2025

@uberduck @superball (but also just fyi...you're right that the planet can't support the lifestyles of these rich assholes, but the vast majority of humans don't live like that or even WANT to live like that, and shouldn't be blamed for the climate crisis)

kauer Jul 24, 2025

@sofiav @uberduck @superball Here is that link to the study showing that people don't get as much as they think they do out of AI:
https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/

Very small sample size though.

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity

Roger Moore Aug 2, 2025

@uberduck @superball
They aren't rushing to AGI; they're rushing to IPO. The hype to product level for AI is as high as the dot com bubble, and I predict it will pop at least as spectacularly.

Simon Brooke Jul 24, 2025

@superball @uberduck if they considered the feedback, they would have to admit -- to themselves, and worse, to their investors -- that their multi-billion dollar investment was a total washout.

So they won't.

#GenAI
#StochasticMonkeysWithTypewriters

Falling forward 🍉Jul 24, 2025

@uberduck
> I am now being required by my day job to use an AI assistant to write code. I have also been informed that my usage of AI assistants will be monitored and decisions about my career will be based on those metrics.

Wow, that's total bullshit. 😡 So sorry you have to deal with that. What an awful place to work.

The Uberduck Jul 24, 2025

@AnnyJoe it's a Fortune 100 company. Talking to my peers, most are like this. Some are far worse.

Ginny Bee (she/her)Jul 24, 2025

@uberduck Lord almighty. this shit makes Dilbert look like Family Circus.

in my Programming 101 course last year one HW assignment was write code w/ ChatGPT. Herding is exactly the word. All i asked it to build was a Rochambeau game; it still took me 3-4 tries to get it to produce working code.

which makes sense, considering that this "intelligence" is literally _hallucinogenic_ automatron data manipulation as a service. IMHO worst waterfall product development result ever 💅🏼

drhoopoe Jul 24, 2025

@uberduck I'm not a coder, but as someone who largely writes academic prose for a living, the thing about using the execution of basic writing tasks as downtime to think about the next problem, and the value of internalizing an argument by executing it "by hand," make perfect sense to me. Being forced to work with a "tool" like that sounds like an absolute fucking nightmare.

William Pietri Jul 24, 2025

@drhoopoe @uberduck For sure. Coding and serious writing have in common that you get much better results if you can hold the whole thing in your head, and at least for me, that requires dwelling in the small details.

I very often have the experience that however much I plan or research in advance, it's only in the act of writing that some of the pennies drop. That I see more deeply, or find a better way through.

The notion that we can automate that mental work strikes me as managerialist foolishness. Wishful thinking from people who often act without real understanding, and think everybody else should too.

Dagnabbit, Pascaline! 💥Jul 24, 2025

All of this!

And having to check if it is correct takes So_Much_Time. Plus it means having to switch heads to try to understand why it would do a particular thing that might not even be what's really needed.

Your situation made me sigh, and I fully understand going home early. The fact they force you to use something that doesn't even care about the company is telling. All for money 😭 While people like you really care and think of what's actually needed.

@drhoopoe @uberduck

Hannah Steenbock Jul 24, 2025

Oh, my goodness, this sounds like an utter nightmare.

Typical, stupid micromanaging by people who do not even want to know how their experts manage their jobs... making their life so much harder because they fell for a hype.

I hope you can find an environment that lets you work in peace and without AI interference.

Paul SomeoneElse Jul 24, 2025

I quit my high paying job mostly because I was pigeonholed into developing with
OpenAI/LLM/agents RAG and all of that. I was even good at it, but it is all so stupid to me.
Across the company they were starting to make us use copilot.

It got to the point that writing good code and developing cohesive systems
was not even on the radar. It was all just vibe code BS. I couldn't do it.

I guess I'm a luddite, but I don't care. I am one of those people that
thinks deleting code is usually of higher value than adding code.
I've always worked with mostly programmers that kinda just went along,
but usually the lead programmers were at least decent to good.
Not anymore at the job I was at. It was all hype, BS, and garbage code.

I worked for a while doing maintenance kind of work on large java projects.
I know how painful it is to fix code where people just added and added and added.
It's like the US tax code.

Anyways good luck :P

The Uberduck Jul 24, 2025

@pkw If I were in a position to quit, it'd be a lot easier to deal with.

I wouldn't call you a Luddite. I'm actively trying to reduce tech debt (in fact, that's my current assigned task), but the "you gotta use AI" edict is company-wide. So I'm writing code to analyze code to convert it from NotInventedHereLang into something that we can actually put on a job description.

You know, assuming we ever hire anyone.

Violet Madder Jul 25, 2025

Luddites are specifically against tech that is unhelpful and exploitative, not tech in general

@violetmadder @uberduck @pkw

I’m a Luddite,
They’re a Luddite
Wouldn’t you like to be a Luddite, too?

https://thenib.com/im-a-luddite/

I’m a Luddite (and So Can You!) | The Nib

What the Luddites can teach us about resisting an automated future.

The Nib

𝐉𝐨𝐧 𝐁𝐞𝐫𝐠𝐞𝐫 Aug 2, 2025

@DavidM_yeg @violetmadder @uberduck @pkw Chant no more your old rhymes about bold Robin Hood...

https://www.dropbox.com/scl/fi/x44bvev3zt6431l6uushm/General-Ludd.mp3?rlkey=ys5uvikcpm22jc96ljk7za86u&st=994mhuyq&dl=0

General Ludd.mp3

Shared with Dropbox

Dropbox

Toni Aittoniemi Jul 24, 2025

@pkw @uberduck This! Deleting code on legacy systems tends be of higher value in the long run than adding.

Illusions of progress now -- leading into technical debt -- will just crash & burn later.

The Frog Jul 24, 2025

"I guess I'm a luddite"

The Luddites were right though. They weren't against technology, they against the way technology was used to destroy lives.

Trendar Aug 1, 2025

@Lily_and_frog @pkw @uberduck

I used to be a technologist, but the older I get the more Luddite I become.

MidniteMikeWrites Aug 1, 2025

@Lily_and_frog @pkw @uberduck if it wasn’t for the luddites and those who fought the political power that employers abused to get workers to accept abhorrent working conditions we wouldn’t have the 8 hour work day, the 40 hour work week, cleaner air, etc.

There’s no shame in being a Luddite if a technology is being adapted in a thoughtless, destructive way. You’re an important counterweight.

Cy Jul 25, 2025

It's funny the Luddites were not opposed to new technology. They just protested it being used to only benefit a few rich people, while everyone else gets screwed. For that they got murdered and enslaved, and now centuries later, the same sort of murderers and slavers are like "oh, you're opposed to new (bad) technology? What a Luddite!" They don't even realize the irony, proudly equating themselves with tyrants and thieves, because they didn't bother to learn anything about working class history.

Guess what I'm saying is I feel you, they're out of their gourd for belittling people who reject technology, and I'm sorry they got away with stealing your high paying job with their petty demands. Good luck to you, too...

EndlessMason Jul 24, 2025

@uberduck
If I wanted to explain software ideas to a distracted pedantic asocial jerk I'd have applied for a project manager role, chaaw

�Jul 24, 2025

@uberduck sounds like an awesome start for maliciois compliance? i would love to see the things break apart and the fall of their product. yes, i am a very bad employee

Frank Heijkamp Jul 24, 2025

@uberduck Vibe coding is throwing a lot of shit at a wall and see if something sticks or not. That's not how professionals should code. You spend more time cleaning things up and restructuring things to create maintainable code that will be not too much of a headache in the (near) future. That future extension could be in two years or two weeks. Write proper code from the start and save time in the long run.

Take care of your mental health.

Kiwï Jul 24, 2025

It's like watching over a child with too much caffeine and crayons while you're trying to finish a painting.

Johns Jul 24, 2025

@uberduck I would just say that my ethics forbid me to follow those instructions and they are free to fire me.
But I'm guessing you are from the US where you don't really have worker protection, it sucks that they force you to use a resource hungry planet ruining technology.

The Uberduck Jul 24, 2025

@Johns_priv There's one state in the country where they have to make up a reason to fire you. In all the others, it's just don't let the door hit you on the way out.

Jared White (ResistanceNet ✊)Jul 24, 2025

@uberduck I am genuinely sorry this happened to you. Wishing you all the best in the face of this relentless onslaught 😖

FWIW, I think I've just lost a client on very similar grounds. They never had any problem with the quality of my software engineering, but suddenly they don't want to work with people who aren't enthusiastically embracing this stuff. It truly makes my head spin. 😭

me·ta·phil, der Jul 24, 2025

@uberduck wow what a hellhole, i'm really sorry for you. what you describe reminds me of this recent self-declared shitpost: https://unstable.systems/@sop/114898566686215926

sofia (@[email protected])

Attached: 1 image A shitpost that's been brewing in my head for a while now

unstable.systems Mastodon (Glitch Edition)

The Uberduck Jul 24, 2025

@metaphil Oh, that's fantastic!

Ernst Gucker Jul 24, 2025

@metaphil @uberduck

So true

PPN Jul 24, 2025

@uberduck company I work for has been recently experimenting with providing Copilot to devs and I must say that I concur. Its aggressive suggestion is annoying, sometimes disruptive and half the time surprisingly difficult to dismiss without wasting time. The chatbot is alright at detecting stupid mistakes I make when I am tired right before the day ends, or at generating obvious tests scaffolding. Otherwise it’s been outputting a lot of garbage and terrible at understanding context

Saule, anar en Mousses Jul 24, 2025

@uberduck maybe this will help you argue : https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity

The Uberduck Jul 24, 2025

@PeryleneBleu If only logic worked on these people.

armadillosoft Jul 24, 2025

{{{ hugs }}}

meek cynics Jul 24, 2025

@uberduck, just write an AI to automate it bro, IMHO piping yes or cat /dev/random into the prompt should be sufficient.

Toni Aittoniemi Jul 24, 2025

@cnx @uberduck I think we're going to need a lot of these hacks! "boss keys" are so back!

Ángela Stella Matutina Jul 24, 2025

They're driving good, competent people like you to burnout.

If we keep going this way without giving a fight, our profession deserves to disappear.

Thank you for your service.