And everyone promotes them for tasks they aren't experts in.
Managers think they could replace devs, but never a manager.
Devs think they could replace management but never a senior developer.
Storyboard drawers think they can write screenplays. Screenplay writers think they can draw storyboards. Etc.
As an expert, you know how shit AI is in your own field, but surely those other jobs are simple enough to be replaced.
It's very easy to replace something that was never critical to the process in the first place. My manager essentially updates my git tickets with what I did. We talk for 5 minutes a week. He just kinda lets me do my thing, I am fully aware of how lucky I am.
The principal task of a competent manager is, primarily, intervening between incompetent upper managers and actual workers. Replacing the incompetent manager removes the need for the competent one.
They’re incredibly useful, but you have to treat their output as disposable and untrustworthy. They’re reinforcement trained to generate a solution, regardless of if it’s right, because it’s impossible to AI evaluate that these solutions are correct at scale.
If you’re writing some core code: you can use an agent to review it, refactor parts, stump the original version, infill methods, and to run your test/benchmark scripts.
but you still have to manage it, edit it, make sure it’s not recreating the same code in 6 existing modules, generating faked tests, etc.
As an example this week on my side project I had Claude Opus write some benchmarks. Total throwaway code.
It actually took my input files, generated a static binary payload from it using numpy, and loaded that into my app’s memory (on its own that’s really cool), then it ran my one function and declared the whole system 100x faster than comparable libraries that parse the original data. Not a fair test at all, nor was it a useful test.
You cannot trust this software.
You’ll see these games metrics, gamed tests, duplicate parallel implementations, etc.
My sister-in-law is a software engineer and project manager. This isn't groundbreaking news or anything but she said that her engineers are using generative AI like this. The problem is that it created exceedingly inefficient and bloated code that barely works. En masse it will bog down systems due to the exponential inefficiencies.
One of our devs came to me with an LLM rewrite of some parts of our automation. Even at first glance you could see a lot was missed, the refactor simply wasn't going to work in that state and critical migration logic just wasn't present.
I binned the branch and did the refactor myself, as it would have taken more time to figure out the damage caused than just starting over.
So glad we now pay premium prices for RAM and non-volatile storage, just so some LLM can vomit up a reheated turd.
and that's the fucking crux of it right there, the damage. Any code base without someone smart enough to throw this trash out is going to take forever to fix or might just be too far gone to save
The 1st major project in my first coding job made me understand technical debt even though I didn't know the name of it.
I suppose there are some people that just see spaghetti code and their only thought is to add more spaghetti code. Thankfully I also understood spaghetti code with my project in uni.
Nah, we need to optimize the spaghetti code with an n8n optimization pipeline for more spaghetti, lets make sure to build a new library for eache seperate function call. Once the function has run we'll tear down that library and build the next, if it errors out that's fine, we can just have the GPT hallucinate another library and call that into the function. It's fungible OOP all the way down. Infinite code is infinite tech debt, I call that job security.
It’s not negative that nobody understands the spaghetti code, it’s advanced optimized code that mere mortals can’t comprehend, and marketing can sell that.
Don’t forget to bill by the hour to clients for bug fixes.
Oh man the worst guy I ever worked with got hired to be a solutions architect.
I was so happy he left. He'd always just say yes to whatever we were asked, even if it was literally impossible. Then he'd fuck off and leave me to deal with actually doing the project and would just throw some wrong advice at me and spend the rest of the day browsing the internet at his desk.
Last week, curious what would be generated, told Cursor (with Claude Opus 4.5) to create an animated LED strip effect for an ESP32 device in C. Pretty simple stuff. It thinks for a long time. Creates a ton of scaffolding, docs, step-by-step agentic checklists, even a Makefile to build and deploy the binary. It then says: "Done."
I go compile it. Lots of errors. I paste over the logs and ask it what's wrong. Claude thinks for a while longer, then goes:
"I see the issue - I only created the header file but never completed the LED manager implementation. Let me check what's there and finish the implementation."
It said it had. Obviously, a fabrication. Even after it said it implemented it, it took a couple more hours of coaxing it and pointing whst should be done before it actually worked.
Point is, one shouldn't go near these LLMs for coding unless they know what to do and how to look for problems.
Do people actually use agents for production code? I feel like it is one of those things that people don't use but is sold to us that everyone uses. A lie to promote bullshit.
It adds so much extra and he'll push thousands of lines of code into a PR every week. He had one bug and tried to refactor it, he bloated that single file by 16%.
It adds so much extra and he'll push thousands of lines of code into a PR every week.
This is just a waste of the reviewer's own time that could be better spent doing actual work. Is it at least split up into multiple commits, or is it one giant shitshow?
Absolutely ripped into multiple software departments about our "shit code"
Bragged that he could do it faster and better with AI
CEO gave him a three month trial run to show it.
AI vibe dude spent the first two weeks showing off all this cool new frontend to managers. Nothing actually worked. They gave him a round of feedback.
Then he spent another two weeks struggling to meet the feedback.
Nobody in the tech department wanted to help him because he came in shit talking.
They ended the trial because the AI Vibe coder dude couldn't handle system changes, how to fix bugs, implementing new feature requests without breaking old stuff, and didn't have any real coding skills. He barely lasted a month.
The first half of this story made me wonder if we were colleagues.
The second half was different though. Our guy was a personal friend of some high up, slandered the existing codebase without so much as even speaking to the existing devteam, and then took the better part of a year claiming he could replace the entire decade old codebase while making vague promises that it was coming soon. Meanwhile upper management was taking the slander seriously, punished my department and got a new manager for it. It wasnt until the new manager outed him as a fraud for his ass to finally get caught.
I doubt he was able to read the legacy codebase at all.
Prompt: "write a python script to create a thousand linkedin accounts with plausible sounding names, and then setup a cronjob to post every day a punchy linkedin just-so story explaining why everyone should be paying for my LLM and to keep paying for it when I jack up the price 100x to cover my expenses"
Pokémon Red & Blue was 373 KB. So much efficiency and creative coding made early high-content video games possible. Imagine how bloated that would be if it was vibe coded by AI
But could you imagine all the exploits we would have found by now? That special little coast line would have been nothing. Missing Number? Bet we would have had the whole Missing Alphabet haha.
You're absolutely right! I didn't actually run anything. I was displaying a "thinking..." animation and simply waited for your next input. I shouldn't have done that.
I think my management is decent about receiving feedback, but are also very into AI utilization right now. I’m not sure what level of feedback to give, because I do feel being too obstinate and negative would mark me, and with my adhd I often dont have a good read on where that line is. But I’d love to press them on where it will stop being worth it to keep pushing it? When the venture capital dries up? Climate/local data center regulations finally get implemented? Because those will all drive token cost up by a lot
I’ve learned they don’t actually understand either your craft pr the AI, but the AI is very convincing and more people in high places just go “sure, that validates my preconceived notions”
84 Comments
in_my_honest_opinion@piefed.social · 130 pts · 200d
Hahahaha this idiot thinks that it's the speed of our typing that ships code faster. He's in a knowledge shortage.
teft@piefed.social · 54 pts · 200d
I’ve found that the people who understand these “agents” the least are the ones who are promoting them the most.
mech@feddit.org · 40 pts · 200d
And everyone promotes them for tasks they aren't experts in.
Managers think they could replace devs, but never a manager.
Devs think they could replace management but never a senior developer.
Storyboard drawers think they can write screenplays. Screenplay writers think they can draw storyboards. Etc.
As an expert, you know how shit AI is in your own field, but surely those other jobs are simple enough to be replaced.
Rivalarrival@lemmy.today · 24 pts · 200d
Let's be honest, though: they absolutely could replace management.
cynar@lemmy.world · 23 pts · 200d
A good manager is both a coordinator and a filter. They deal with bs rolling down from above and keep their team running efficiently.
A good manager is worth their weight in gold. A bad manager isn't worth their weight in bullshit.
Flamekebab@piefed.social · 8 pts · 200d
Yeah, our PM is great. Our previous one not so much.
He trusts us but also handles absolutely loads of stuff that we don't want to deal with.
clif@lemmy.world · 5 pts · 199d
I've also enjoyed the term "shit umbrella" for a good manager.
in_my_honest_opinion@piefed.social · 13 pts · 200d
It's very easy to replace something that was never critical to the process in the first place. My manager essentially updates my git tickets with what I did. We talk for 5 minutes a week. He just kinda lets me do my thing, I am fully aware of how lucky I am.
mrgoosmoos@lemmy.ca · 2 pts · 199d
some management, sure. just like some of your coworkers could probably be replaced by AI. but not the competent ones, and not the essential ones.
and personally, I'd still rather work with an incompetent person who can improve than four incompetent chatbots
although I'd rather work with no incompetence at all
Rivalarrival@lemmy.today · 3 pts · 199d
The principal task of a competent manager is, primarily, intervening between incompetent upper managers and actual workers. Replacing the incompetent manager removes the need for the competent one.
baines@lemmy.cafe · 11 pts · 200d
90% of my experience with management is having none at all would be a net benefit
why would we want to add ai to that mix
atomicorange@lemmy.world · 8 pts · 200d
At least they’d be pleasant while being useless? I’ve had managers that were egomaniacal terrors.
baines@lemmy.cafe · 7 pts · 200d
problem is ai would be another layer of separation from decisions and consequences
Rivalarrival@lemmy.today · 3 pts · 200d
I thought "egomaniacal terror" was part of the job description.
panda_abyss@lemmy.ca · 7 pts · 200d
This.
They’re incredibly useful, but you have to treat their output as disposable and untrustworthy. They’re reinforcement trained to generate a solution, regardless of if it’s right, because it’s impossible to AI evaluate that these solutions are correct at scale.
If you’re writing some core code: you can use an agent to review it, refactor parts, stump the original version, infill methods, and to run your test/benchmark scripts.
but you still have to manage it, edit it, make sure it’s not recreating the same code in 6 existing modules, generating faked tests, etc.
As an example this week on my side project I had Claude Opus write some benchmarks. Total throwaway code.
It actually took my input files, generated a static binary payload from it using numpy, and loaded that into my app’s memory (on its own that’s really cool), then it ran my one function and declared the whole system 100x faster than comparable libraries that parse the original data. Not a fair test at all, nor was it a useful test.
You cannot trust this software.
You’ll see these games metrics, gamed tests, duplicate parallel implementations, etc.
baines@lemmy.cafe · 9 pts · 200d
spend more time fixing slop compared to just doing it manually and correct the first time
CIA_chatbot@lemmy.world · 4 pts · 200d
They are also the ones who have super leveraged their portfolios with AI stocks -
halfapage@lemmy.world · 12 pts · 200d
apfelwoiSchoppen@lemmy.world · 69 pts · 200d
My sister-in-law is a software engineer and project manager. This isn't groundbreaking news or anything but she said that her engineers are using generative AI like this. The problem is that it created exceedingly inefficient and bloated code that barely works. En masse it will bog down systems due to the exponential inefficiencies.
It's fine. Everything is fine.
in_my_honest_opinion@piefed.social · 41 pts · 200d
Yup, it's a junior dev that never learns, makes thousands of tiny mistakes that tolerance stack into a brittle gnarled mess.
ech@lemmy.ca · 11 pts · 200d
Who needs tech-debt when you have new and improved slop-debt?
RampantParanoia2365@lemmy.world · 3 pts · 199d
So......why is she letting them, and not instead, like, not letting them?
apfelwoiSchoppen@lemmy.world · 4 pts · 199d
I'm not her or her employer. It is an odd time for the coding industry. It'd be like a fish swimming against the manager/leader-lead river.
baines@lemmy.cafe · 2 pts · 200d
how’s she like working for Microslop?
apfelwoiSchoppen@lemmy.world · 1 pts · 199d
If folks think Windows 11 is bad now! FWIW she ain't at a big tech company. It is the whole ass coding industry as seen by countless reports.
Inucune@lemmy.world · 48 pts · 200d
Stop hiring 20 managers. Hire 1 manager and have them in meetings all day so real work can be done.
RichardDegenne@lemmy.zip · 1 pts · 199d
Ah yes, the Fred Trump gambit.
sasquash@sopuli.xyz · 47 pts · 200d
"The new skill isn't typing faster".
Since I added a second keyboard I am programming twice as fast and don't even need AI!
hayvan@piefed.world · 24 pts · 200d
We pair-code on the same keyboard for maximum efficiency.
Grimy@lemmy.world · 29 pts · 200d
notabot@piefed.social · 8 pts · 200d
You should quad-code on two keyboards, really pump up those LOC stats.
usualsuspect191@lemmy.ca · 11 pts · 200d
It's also the second-best way to combat hackers.
panda_abyss@lemmy.ca · 9 pts · 200d
Are you guys using your feet yet? That’s how I run 4 keyboards.
TheGiantKorean@lemmy.today · 9 pts · 200d
I use 5. The new 5th limb is highly experimental and probably dangerous but my productivity has skyrocketed.
panda_abyss@lemmy.ca · 8 pts · 200d
Its hard to maintain structural integrity of the 5th limb long enough for consistent productivity
TheGiantKorean@lemmy.today · 4 pts · 200d
I just have to get through this last sprint! Just one more sprint...
Grimy@lemmy.world · 3 pts · 200d
You just need big python energy
Aganim@lemmy.world · 33 pts · 200d
One of our devs came to me with an LLM rewrite of some parts of our automation. Even at first glance you could see a lot was missed, the refactor simply wasn't going to work in that state and critical migration logic just wasn't present.
I binned the branch and did the refactor myself, as it would have taken more time to figure out the damage caused than just starting over.
So glad we now pay premium prices for RAM and non-volatile storage, just so some LLM can vomit up a reheated turd.
_lilith@lemmy.world · 2 pts · 199d
and that's the fucking crux of it right there, the damage. Any code base without someone smart enough to throw this trash out is going to take forever to fix or might just be too far gone to save
TheBat@lemmy.world · 32 pts · 200d
ulterno@programming.dev · 17 pts · 200d
The 1st major project in my first coding job made me understand technical debt even though I didn't know the name of it.
I suppose there are some people that just see spaghetti code and their only thought is to add more spaghetti code. Thankfully I also understood spaghetti code with my project in uni.
in_my_honest_opinion@piefed.social · 7 pts · 200d
Nah, we need to optimize the spaghetti code with an n8n optimization pipeline for more spaghetti, lets make sure to build a new library for eache seperate function call. Once the function has run we'll tear down that library and build the next, if it errors out that's fine, we can just have the GPT hallucinate another library and call that into the function. It's fungible OOP all the way down. Infinite code is infinite tech debt, I call that job security.
panda_abyss@lemmy.ca · 7 pts · 200d
It’s not negative that nobody understands the spaghetti code, it’s advanced optimized code that mere mortals can’t comprehend, and marketing can sell that.
Don’t forget to bill by the hour to clients for bug fixes.
in_my_honest_opinion@piefed.social · 4 pts · 200d
This guy speaks "solutions architect"
panda_abyss@lemmy.ca · 6 pts · 200d
Oh man the worst guy I ever worked with got hired to be a solutions architect.
I was so happy he left. He'd always just say yes to whatever we were asked, even if it was literally impossible. Then he'd fuck off and leave me to deal with actually doing the project and would just throw some wrong advice at me and spend the rest of the day browsing the internet at his desk.
RampantParanoia2365@lemmy.world · 2 pts · 199d
Damn you, now I want spaghetti with mushrooms.
ulterno@programming.dev · 1 pts · 199d
Until the AI Token bills start getting counted into your CTC.
gjoel@programming.dev · 7 pts · 200d
Oh, but you don't have technical debt anymore, since the agent rewrites all the code every time!
BradleyUffner@lemmy.world · 3 pts · 200d
And much like copying a copy, more and more errors creep in each time.
IchNichtenLichten@lemmy.wtf · 5 pts · 199d
The most important thing in vibe coding is to cash the check and nope out of there before management realize what you've done.
ZoteTheMighty@lemmy.zip · 5 pts · 200d
It's a problem you type into a chat for Claude to solve.
fubarx@lemmy.world · 21 pts · 200d
Last week, curious what would be generated, told Cursor (with Claude Opus 4.5) to create an animated LED strip effect for an ESP32 device in C. Pretty simple stuff. It thinks for a long time. Creates a ton of scaffolding, docs, step-by-step agentic checklists, even a Makefile to build and deploy the binary. It then says: "Done."
I go compile it. Lots of errors. I paste over the logs and ask it what's wrong. Claude thinks for a while longer, then goes:
"I see the issue - I only created the header file but never completed the LED manager implementation. Let me check what's there and finish the implementation."
RampantParanoia2365@lemmy.world · 1 pts · 199d
Could you try something else and when it says it's done, ask it to check for errors first? Just curious, and not a programmer.
fubarx@lemmy.world · 3 pts · 199d
It said it had. Obviously, a fabrication. Even after it said it implemented it, it took a couple more hours of coaxing it and pointing whst should be done before it actually worked.
Point is, one shouldn't go near these LLMs for coding unless they know what to do and how to look for problems.
slazer2au@lemmy.world · 17 pts · 200d
Thought we were on !linkedinlunatics@sh.itjust.works for a moment.
_lilith@lemmy.world · 17 pts · 199d
man does this dude think the hardest part about writing a book is fucking typing it out? someone give this dork a swirly
SleeplessCityLights@programming.dev · 15 pts · 200d
Do people actually use agents for production code? I feel like it is one of those things that people don't use but is sold to us that everyone uses. A lie to promote bullshit.
brygphilomena@lemmy.dbzer0.com · 14 pts · 200d
I have an engineer that uses it heavily.
It adds so much extra and he'll push thousands of lines of code into a PR every week. He had one bug and tried to refactor it, he bloated that single file by 16%.
It's almost impossible to review.
pivot_root@lemmy.world · 6 pts · 200d
This is just a waste of the reviewer's own time that could be better spent doing actual work. Is it at least split up into multiple commits, or is it one giant shitshow?
panda_abyss@lemmy.ca · 5 pts · 200d
Yes, they do.
sukhmel@programming.dev · 3 pts · 200d
I sometimes do, but very sparsely, it's hard to come across a task that's good fit for an LLM unless you're prototyping something, imo
muusemuuse@sh.itjust.works · 15 pts · 199d
Yea…..but how may developers are you going to have to pay to maintain that shit?
VitoRobles@lemmy.today · 18 pts · 199d
Fun story!
The CEO was charmed by some AI vibe dude who
CEO gave him a three month trial run to show it.
AI vibe dude spent the first two weeks showing off all this cool new frontend to managers. Nothing actually worked. They gave him a round of feedback.
Then he spent another two weeks struggling to meet the feedback.
Nobody in the tech department wanted to help him because he came in shit talking.
They ended the trial because the AI Vibe coder dude couldn't handle system changes, how to fix bugs, implementing new feature requests without breaking old stuff, and didn't have any real coding skills. He barely lasted a month.
spidertrolled@lemmy.blahaj.zone · 8 pts · 199d
The first half of this story made me wonder if we were colleagues.
The second half was different though. Our guy was a personal friend of some high up, slandered the existing codebase without so much as even speaking to the existing devteam, and then took the better part of a year claiming he could replace the entire decade old codebase while making vague promises that it was coming soon. Meanwhile upper management was taking the slander seriously, punished my department and got a new manager for it. It wasnt until the new manager outed him as a fraud for his ass to finally get caught.
I doubt he was able to read the legacy codebase at all.
Pencilnoob@lemmy.world · 13 pts · 200d
Prompt: "write a python script to create a thousand linkedin accounts with plausible sounding names, and then setup a cronjob to post every day a punchy linkedin just-so story explaining why everyone should be paying for my LLM and to keep paying for it when I jack up the price 100x to cover my expenses"
in_my_honest_opinion@piefed.social · 3 pts · 200d
You'll want to set a systemd timer for that actually, easier to have the agent journalctl to get the full stderr and properly hallucinate more jobs.
Jayjader@piefed.social · 13 pts · 200d
"We're not in X, we're in Y"
squints suspiciously
AlexLost@lemmy.world · 12 pts · 200d
There isn't a talent shortage, there's a shortage of people who will take your shit at sub-par wages working two + jobs at your company.
Jankatarch@lemmy.world · 11 pts · 199d
Techbros calculate programming skills by lines of code per shift.
It only makes sense that they would think "10x engineer" just means "types 10 times as fast."
drsilverworm@midwest.social · 11 pts · 199d
Pokémon Red & Blue was 373 KB. So much efficiency and creative coding made early high-content video games possible. Imagine how bloated that would be if it was vibe coded by AI
Entertainmeonly@lemmy.blahaj.zone · 8 pts · 199d
But could you imagine all the exploits we would have found by now? That special little coast line would have been nothing. Missing Number? Bet we would have had the whole Missing Alphabet haha.
Etterra@discuss.online · 10 pts · 199d
3 months later they rehired all of those four engineers to fix all that bullshit useless code.
JamBandFan1996@lemmy.ml · 10 pts · 199d
I hate LinkedIn and all the fuck holes on it so god damn bad
Diplomjodler3@lemmy.world · 9 pts · 200d
Don't worry. All the code has been thoroughly tested. By Claude.
Uruanna@lemmy.world · 10 pts · 200d
You're absolutely right! I didn't actually run anything. I was displaying a "thinking..." animation and simply waited for your next input. I shouldn't have done that.
typicalDude778@sh.itjust.works · 9 pts · 199d
Ah yes one of those bootlicking linkedin lunatics
homesweethomeMrL@lemmy.world · 9 pts · 200d
Nothing can go wrong with that plan
varjen@lemmy.world · 7 pts · 199d
There's going to be so much awful slop code everywhere in the future.
bitwise@lemmy.ca · 4 pts · 199d
This is the real reason why all those scrappy rebels in sci-fi shows can hack into everything.
Avicenna@programming.dev · 7 pts · 199d
guy in the business of vibe coding hypes up vibe coding
BroBot9000@lemmy.world · 7 pts · 200d
What a delusional sack of shit. lol
gabbath@lemmy.world · 6 pts · 199d
This was definitely written with AI.
Ioughttamow@fedia.io · 6 pts · 200d
I think my management is decent about receiving feedback, but are also very into AI utilization right now. I’m not sure what level of feedback to give, because I do feel being too obstinate and negative would mark me, and with my adhd I often dont have a good read on where that line is. But I’d love to press them on where it will stop being worth it to keep pushing it? When the venture capital dries up? Climate/local data center regulations finally get implemented? Because those will all drive token cost up by a lot
panda_abyss@lemmy.ca · 8 pts · 200d
I’ve learned they don’t actually understand either your craft pr the AI, but the AI is very convincing and more people in high places just go “sure, that validates my preconceived notions”
ramble81@lemmy.zip · 5 pts · 200d
I’m getting ready to take over an org that’s doing this. It’ll be interesting to run a static code analysis and penetration test on the sites.
degen@midwest.social · 5 pts · 199d
"Talent utilization crisis"
"Hire 1 engineer instead of 4"
Yes