this post was submitted on 10 Sep 2026
280 points (90.9% liked)
Technology
88725 readers
2929 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I haven't dug into the code, but blanket hate for AI in code is extremely frustrating. There's a difference between vibe coding and the hundred other ways to "use AI" on a project.
It's funny because anyone seeing this as black and white clearly hasn't ever opened an IDE.
I've been a programmer for two thirds of my life up to this point. I don't see how the machine that generates code that must be assumed to be broken, and statements about said code that must be assumed to be wrong, until both have been manually verified, because the "hallucination" issue is inherent to the design of LLMs and cannot be solved, is supposed to make my life easier. Especially since a well known problem in computer science is that writing code is more fun than reading it.
If you're using AI for code generation, I hope you like reading code.
If you're using AI to learn your way around an existing codebase, I hope you're prepared to fact check everything it tells you.
If you're using AI for code review before merging a PR, fucking yikes. What made you think the sycophant machine that once told users a soggy cereal bar was a great business opportunity and cannot even transcribe McDonald's orders correctly was suitable for that purpose?
If you're telling yourself it's okay because there's a human in the loop, please read some studies about attention fatigue and consider that increasing the number of critical decisions a person must make per day, while giving them a suggested answer that's right 90% of the time and catastrophically wrong the remaining 10%, and the only way to tell the difference is to read in depth, might not have the desired effect. Especially if the human in question is then put under time pressure.
And all that is before I go on a five paragraph rant about the ethical transgressions of every major AI company.
It's pretty black and white for me.
This in spades.
To add; now there's an issue in the generated code that's been pushed. What to do? Read the code? Have new code generated? Abandon the project?
Abandon the project
quoted for emphasis. you have more patience than I do in telling sycophants why their take is idiotic.
The fact that it isn't perfect is the grey. And sure, it has a lot of problems but not only is it still useful anyways, but those problems are being ironed out. They probably won't disappear completely but the tech is already miles further than what we had 2 years ago.
To make you understand what I mean by black and white, you have a user that put "AI is evil" in big bold letters in the comments. The hate is almost cult like for part of the population, and imo, it's mostly because the media tells them to be mad.
Most of it depends on the task in any case. I don't work with massive code bases and it doesn't take long reading through a few scripts, running it and checking the output. The people I know that do work with huge code bases are using it as well, but I can't really talk for them either. They don't hate it tho.
The person who put "AI is evil" in big block letters is objectively correct. See homes getting claimed under eminent domain to build power lines for datacenters, electricity prices for consumers rising as service quality gets worse, and water usage to the tune of millions of gallons per datacenter per day. See companies buying rare books in bulk, cutting them apart so they can be scanned faster, and not publishing the scans, literally destroying culture so they can have one more scrap of training data guaranteed to be free of slop that their competitors don't have. See companies operating web scrapers that deliberately evade every block that website owners try to use to stop them from reading their sites, and then scraping so many different pages so frequently that it's indistinguishable from a DDoS attack, hosting bills skyrocket, servers crash, indie webmasters pay the price for corporate greed. See xAI operating natural gas turbines on the property of their datacenter for extra electricity in flagrant violation of the Clean Air Act, spewing toxic smoke into the air and generating powerful, inaudible infrasound vibrations that have been linked to everything from nausea to irritability to hearing loss to seizures. See Anthropic treating Minority Report as an instruction manual. See how ChatGPT use affects test scores. See AI psychosis. See chatbot-induced suicides. See deskilling. See how little AI companies care.
Yeah. Even if you run the model entirely on your own 100% solar-powered hardware, there's still the problem of training data, and there's still the problem of the myriad adverse effects on the user.
Sorry, I don't bootlick the copyright machine. Anyone with half a brain can see it's transformative. I don't get why collage is fine but AI isn't, it's just copyright mega corps and data brokers that smell blood in the water. All the data isn't even owned by the artist, you are fighting for universal records, not your local guitarist.
For the climate aspect, it's mostly grid issues and corruption. It's laughable to blame AI when we have president pedo prostrating himself in front of oil companies and destroying renewable projects. We need regulations and all that on where they can be placed and how they get powered, but in the end, it's just software. I could use AI 8 hours a day for my whole life and I still wouldn't be close to the damage one ticket to Europe does. It's a scape goat being served by a corrupt pro oil lobby.
Don't get me wrong, I know what datacenters entail with the amount of people they serve, but if you look at it at an individual level, it's a drop in the bucket, or it would be if they weren't using plane turbines. We need regulations to curb the stupidity, but it's doesn't make it all inherently bad.
I don't think the copyright thing is an issue.
Climate wise, it's only an issue because of a corrupt administration, it's not a death blow to it. It's at the heart of it overblown by a reasonable measure imo.
Not the person you are replying to, but if there is no copyright there is no IP "theft".
And the fact how it is used in practice, actually, in reality, is the very dark grey that some people call "black" for better clarity. The fact that you can use a large butchers knife for practical things does not mean that entering a primary school with one is the right thing. What are the actual intentions for grabbing that tool?
These are supply chain attacks and in the case of xz utils, the attacker had gone to extreme lenghts to hide the attack from a well-meaning, good-hearthed but overworked and burnt out solo maintainer.
To compare this to bugs that people unwittingly introduce in normal human-written code is not sincere.
Okay, but they still happen with orders of magnitude less frequency than bugs in AI code. Consider that the time between when rsync first adopted LLM-generated code and users en masse reporting rsync internal protocol errors during a backup was on the order of months.
The author of that article admits that the sample size is not large enough to draw meaningful conclusions.
But besides that, I believe LLM code generators can be a useful tool, provided you are willing to go over their output with a fine-tooth comb and assume it is broken until you have proven otherwise, because the hallucination problem is inherent to the technology and they're never going to completely solve it, and are willing to overlook the myriad ethical issues with all major LLMs in existence today.
I'd really like to see a source for the first guy's numbers, especially how he accounts for things like training and water usage at the power plant, and as for the second guy, Anthropic is not preserving shit. The Internet Archive preserves books. Anthropic scans them and doesn't publish the scans so that they can train an LLM that might or might not be able to regurgitate some fragments of that text, and publish that. That's preservation in the same way that painting a picture of you is keeping you alive forever.
Also, neither of those address the effects on creatives' livelihoods or the mental health of LLM users. Chatbot psychosis is real. People who routinely use LLMs to do things provably become worse at doing them without them. Students use LLMs to make an end run around having to learn everything they need to know to be effective members of a society, like how to articulate their points, how not to fall for rhetorical traps, what history was really like, and between that and the disastrous effects of No Child Left Behind, teachers are quitting in droves and there's a literacy crisis.
I froze my rsync package the day that was announced. fuck the dev :(
I understand this but, you have to consider the ecological impact of it, the thousands or more people whose work was stealed to train AI, the constant shoving of AI into everything and the problems that have arise from it, the campain AI companies did and do telling people they are going to lose their jobs and constant fearmongering, the constant blatantly lying of "AGI achieved", the layoffs by senile CEOs, the debt that is causing because stupid people in power gonna be stupid, etc. There are plenty of reasons to hate AI and what it is doing to our world and people lives in general. Sure there are positives but the negatives clearly outweighs the positives.
On top of that, regardless of however you feel about AI, there are so many questions about copyright and ownership here, that as-of-yet remain untried and unanswered by the courts, that it's essentially impossible to know what the legal ramifications will turn out to be, when accepting "AI code" in your project. I definitely fear that it could turn out to be a sort of trojan horse against especially the GPL, and other copyleft licenses. I conaider it to be risky to use AI code, on that basis alone.
True, but how will you prove that code was written by AI and not yourself? Any people can generate code, modify some variables to incorporate in their code, and it won't be any different than copying it from stackoverflow. Unless the code is written in a "very specific AI way" it won't be possible to distinguish.
Oh, I don't think any court would rule that using LLM generated code automatically infringes on the copyright of everyone whose text was used to train it, independently of the degree of similarity between the generated output and the original text(s).
But LLMs have a habit of sometime, and not too infrequently either, reproducing parts of their training material verbatim, as their output. As long as it's only sufficiently small "snippets" of text or code, it's unlikely to raise significant copyright concerns - but the fact is that it's not actually that rare that an LLM will replicate larger sections of text from its training material, spanning across many lines. What's more, LLMs don't advise you of this, when their output is a replica of a significant part of an already-existing, copyrighted text. But you'll be infringing on somebody's copyright, whether you know it or not, irregardless of whether you aquired the text from an LLM or elsewise. That's a big liability to accept in a large project, that frequently uses LLM-generated code. It would be a minefield.
I'd hope that they will stick to non-retroactivity for new laws and rulings on AI, or else the world will have a giant mess on its hands. But I have the feeling that we're moving towards an eventual, enduring weakening of copyright law, which is totally fine by me. Copyright has done more harm than good to the world (imo).
Usually, if it's noticeable that AI was used on code, the code is flawed or, at least, has copyright concerns.
Agreed.
I strongly remember a CompSci professor ~2005 being disgusted with auto-generated Java GUI. And yes it was clunky and more memory intensive than it could have been, but also, I really don't care about how GridView is defined most of the time.
Bad code is bad code, touched by AI or not. For working code, a volunteer group spending resources to determine if it might have been touched by AI is at the very least an incredibly frustrating standard to try to meet. And very likely impossible to verify in any meaningful way.
I am not talking about the Linux kernel or the Rust project - they have stabdards way above average commercial projects - but generally "using ai" in practice means to use it for code generation. Then, programmers are expected to review the output but are not given extra time for it. This creates an extremely slippery slope leading to mostly unsupervised code creation. Also because reading and understanding buggy third party code means more effort than writing new code.
Then there is the idea to get rid of the bugs by "debugging", but debugging is frankly not a software engineering strategy, because you can always just recognize a fraction of the bugs that your process created in the first place. Debugging is also more expensive tha writing good code.
Then, what makes the process between product managers and software developers very often a bit friction-loaded is that good developers always want to have clear specifications what they are supposed to build.
LLM code generation largely replaces clear specifications with plausible assumptions. That appears to make the process fast, but the result is often wrong. At times obviously wrong, and at times subtly and catastrophically wrong.
Excellent observation, which aligns with my observation that fans of "coding assistants" can't wrap their head around algorithms, not to mention a sound software architecture.
Eh, I now have witnessed two cases where I asked a software team lead of a project I was working for for a precise interface or API for a planned function and they looked at me as if I had talked about going on a giraffe ride with a bunch of alien Martian schoolkids.
(And subsystem interfaces is really a subject I was taught in 10th grade by a math professor who himself had been a student of Niklaus Wirth. Maybe if you stand on the shoulders of giants, you should not be pissing on them...)
It was already difficult to find people who know how to phrase good requirements and / or design an architecture before the slop machines got introduced. It will be so much worse for the next couple of decades, with idiots bullshitting their way past HR into a position.
you misspelled "extremely deserved"
Look at the comments here. Do you think the haters here care?
Shh, you’re using logic on social media.
If the haters only realized that literally every software company has been using it since at least last fall, we’d have a helluva lot more Amish.
Noooo All AI is Baaad!!1!
/s
It's tiresome.