LLMs suck at writing code, but they are absolutely amazing at reading code and finding bugs. I am firmly against the usage of LLMs to write anything, but they have a place as a diagnostics tool when verified by a human.
not to be an AI bootlicker, but i don’t get why people treat AI-generated code like it’s the Plague 2. it’s not; all code is constrained by the programming language, and so it either implements the feature correctly… or it doesn’t. yes, human review is undeniably important, but other than that, AI-generated code and human code is still all logic at the end of the day
because it’s a perfect storm. The way it’s being managed by businesses is causing :
- Layoffs
- RAM shortages
- Propaganda
- Environmental problems
So, the tool is getting a bad rep. never mind that for over a decade, machine learning was something that was tinkered away and thought of as “one day this is going to be amazing”, now it’s being wielded by assholes trying to be TRILLIONAIRES and people are losing their jobs, access to technology, political and environmental collapse.
To top it off, before you could contextualise ML in statistics and say “it’s a work in progress, some things are not quite right” but now you have ads saying “Replace your humans with our AI/Robots” as if it’s 100% reliable, a drop in replacement and not going to provide any social ills.
Exactly. The concept of an LLM is perfectly fine, we’ve had its precursors like markov chains and neural nets for decades, the issue is how their being created and used these days, particularly the scale of it. LLMs are the shiny new toy the way blockchains were previously, so all the tech bros and utterly clueless CEOs are rushing out to embrace the hot new thing and demand it be shoved into every nook and crany because they think if they don’t they might not make every last cent they possibly might.
Oh. No. I’m sorry, unrelated to the ai debate, but I have to correct this. Software architecture is intensely not raw logic sequences like you learn in uni. Everything before compilation is cooperative abstract modeling that does not necessarily impose any change at all in the binary. You are essentially teaching laymen that all programming languages for the same plattform are a creative choice, that extensibility, robustness, scaling or maintenance are just completely imagined fields of research. I’m sorry but that statement is complete nonsense misinformation. The binary itself, while it’s a stretch; sure, in a way it can be considered only logic. From a certain angle. But that viewpoint is actively unhelpful for understanding what code is. The product is always so much more than the machine code. Consider writing raw arm instructions vs the obfuscated js that our engines churns through that decides most of our executing logic. All that is architecture and affects output only through the lens of the machine code. You have to enter the "code“ any time anything breaks, or is updated, if even a single thing is not as expected, or if you wanted to iterate on the program, or collaborate, you have to move into abstract land and then the architecture matters. The term spaghetti code does not mean that the logic is messy. The compiler wil usually remove most such mistakes. Instead the whole architecture, the “blueprint” the “workbench” is the magic of code, even if it is often obscure for the user is much more important for their experience than the raw logic.
I think the people most extreme about it haven’t done much development.
Or they have, professionally, and associate it with very real AI psychosis at work.
Or they’re in school and see LLMs ruining the student body, as we’ve seen some in other Lemmy posts.
It all comes down to ethical tool sourcing, and using those tools in moderation, responsibly. Like all sorts of other automation/technology that’s popped up in programming…
But I guess I’m saying I can sympathize with absolute “no-LLM-code” positions, too. The way true vibe-coding has been peddled to the public really has done tremendous damage, and honestly the space deserves the negative attention its getting, and then some.
The fact of the matter is as an AI hater I do actually trust AI generated code if run through the right people. I just don’t know when the people I trust will trust the AI more than themselves, and that’s part of the worry. Are they going to be as diligent on checking their AIs bug fix #125 if the first 124 were perfect? Maybe, maybe not. I certainly am not trusting AI code in the hands of vibe coders who have no idea what they’re doing and are only contributing to a project because it happens to allow AI contributions.
I also just don’t like LLMs in general, the way they collected data, the way they were trained on data, the way they are affecting communities across the globe, and the harm the people at the top of these companies are doing, not to mention the job loss which is really not a main concern of mine. I don’t like the idea of people using tools that are so harmful if it can be avoided.
I’d wager people like linux devs will learn exactly how unreliable they are, quickly, once they start using them. Its kind of impossible to miss.
There’s a chance some devs succumb to AI psychosis, but that’s why these large open source projects are institutionalized. No one person can ruin them, by design.
As for ethics, I think the open weights LLMs are far more ethical, albeit far from perfect. They’re trained modestly, presented as tools, apache licensed, self host able, more efficient to run, and good enough; people just need to know they exist as an alternative.
Open weight LLMs are not trained any differently than all the others. They just publish the weights. It’s still massive theft run through a data center.
There is no (good) coding LLM trained ethically because securing rights to the amount of code and text required has not been attempted and is likely not possible. The best you can do is a company who hid a “by using this service you give us rights to use your code/writing to train our LLM” into a six page EULA no one reads and that’s just a different kind of unethical.
Some models like Nemotron use open datasets, and utilize hardware more efficiently than OpenAI and such.
Its (IMO) also quite different when the weights are Apache licensed. If absolutely zero profit is involved, that gets close to the “fair use” umbrella, IMO, like other noncommercial derivative works of restrictively licensed content (fanfics/fanart).
You are not wrong about the closed datasets though, for the vast majority of LLMs. Lord knows what the Chinese trainers or OpenAI use to train…
But anything truly “open weights” is still a different order of magnitude in the ethics scale, IMO. I view the primary issue to be profiting off the scraping and training; take that away, and its not nearly as egregious.
ethical tool sourcing
Almost everybody that brings up ethical LLMs is full of shit because they point to them existing as a defense then turn around and use claude.
I don’t. I’ve never had an OpenAI or Claude subscription. Basically the only thing I’ve used Claude API for is science/testing when comparing it to an open weights model, and I haven’t done that in some time.
I’ve never felt a need to, either.
Not that Im perfectly innocent or anything. Or that you’re wrong. People DO exactly this, and it makes me scratch my head.
Antoine’s argument is still very naive and short sighted.
Only because some corporations are linked to fascism, does not make the technology inherently fascist.
If these links are an issue, we should also all stop using Microsoft and Google, because they offer Isreael computing power, and thus every service is now supporting genocide. So if you are that user, all fine, enjoy an off-grid life. But don’t pretend to be better, if you still rely on those companies in the first place.
I agree that the corpos being linked to fascism doesn’t make the technology fascist, but plenty of people are switching from MS and Google for that among other reasons. Acting like any interaction with these companies is just as bad as going all in is completely incorrect. All small acts of resistance add up and are important, even if you can’t totally avoid a company that is causing harm. Taking a bus is not as harmful as driving a car just because they both use fossil fuels.
there are robust anti Microsoft and anti Google movements getting more and more mainstream over time. there are plenty of great alternatives to MS and Google software and services. Linux and FOSS as a whole is benefiting immensely from anti Microsoft sentiment right now. We can and should boycott these two terrible companies
“Avoiding fascist-supporting companies is inconvenient” is a poor excuse, especially when it’s so easy to degoogle compared to 10 years ago.
NetBSD banned submitting “Tainted Code” in their commit guidelines…
https://www.netbsd.org/developers/commit-guidelines.html
May be time to dump Linux, Linus.
I’ve been thinking about doing just this, but I haven’t seen any evidence that BSD is actually usable in real life yet.
Linux is actually good today, it’s a viable alternative to the big corpos for virtually anybody, and the community is big enough to be sustainable. Also Steam has made Linux gaming a real thing.
If I could build a little pc to try BSD out on I would, I’d love for the competition to be there… But even a raspberry pi is like $500 today. It’s insane. Fucking AI.
Still looks like they allow it as long as it gets approval by the core team. Not saying I agree with that choice, but there’s still a very real chance LLM-generated code makes it into NetBSD.
Code generated by a large language model or similar technology, such as GitHub/Microsoft’s Copilot, OpenAI’s ChatGPT, or Facebook/Meta’s Code Llama, is presumed to be tainted code, and must not be committed without prior written approval by core.
Oh, please… is it still a text generator, or is it already a religion?
Didn’t you love T9 back in the days?
I’m just a movement generator.
Saying “it’s a text generator” cannot be the full force of your criticism, if the text it generates is incredibly useful. What could AI be on a screen if not a text generator?
Your views have got to be formed by WHAT text is generated given that AI cannot express itself through any other channel than text.
If you can give AI a deeply complex code base and ask it to find the likely cause of an obscure bug, and the solution is right, then you’ll have to acknowledge that some form of reasoning is going on, no?
And given that no other system has previously enabled this form of reasoning, it seems a shame to judge AI on semantics, rather than usefulness.
I’m just a movement generator.
Which is way more complex than just predicting text, because you have to predict physics and all that stuff. There is a reason most robots controlled by neural nets have failed hilariously so far.
then you’ll have to acknowledge that some form of reasoning is going on, no?
By that logic you could say a compiler is reasoning. But the reasoning displayed was happening when the compiler (or LLM training material respectively) was written.
It can’t reason. It doesn’t think. It makes shit up as it was trained to do. It can sort of detect patterns and make a guess about what response would get it the highest score, but it’s so error-prone that a human has to verify the output anyway. It’s difficult to tell whether the perceived benefits outweigh the costs. There are many different types of LLMs and use cases.
As a person who values life and our environment, I’d rather not participate in the sloppification of our world. As a Linux user, I am wary of vibe coded flaws slipping under the radar and screwing up my system.
I have to verify the output of what humans generate, too. I’m constantly looking at the source of claims. The AI is no different and it’s not nearly as wrong as a Person is.
If you can give AI a deeply complex code base and ask it to find the likely cause of an obscure bug, and the solution is right, then you’ll have to acknowledge that some form of reasoning is going on, no?
Absolutely not. LLMs are just prediction engines for words.
If a LLM has the right answer, it only means that the model had the correct training data for your issue. Nothing more.
That is absolutely not how it works at all
Uh huh, and how does it work?
What do you mean? That’s exactly how LLMs work. That’s why they “hallucinate”.
No they hallucinate because they lose context but are tasked to answer anyways. They also don’t really lose context as much as they did a year ago.
Their training data has nothing to do with it unless you’re using a local one without web access.
ML is my career and I’ve been doing it for decades. You’re wrong. LLMs hallucinate with full context.
I miss when we thought the future was ML.
Saying “it’s a text generator” cannot be the full force of your criticism,
That is no criticism, just a statement.
My criticism goes to people who assign some metaphysical meaning to AI, or want to treat all it’s output like a contagious pest.
You are on Lemmy so defending AI use will net you down votes. I am an anti-AI person, practically, but you make a very fair and objective point. 👍 I up vote you, not because I agree with AI use but because this is a proper point to make in a proper discussion, which is what this place should be about.
Kudos.
So Debian, what are you going to do now?
Edit: nm, they allow LLM written code now.
In his commit message:
[And this was a debug session from hell, enormously helped by an AI doing much of the grunt-work.
I’d like to call it my tireless helper, but the AI several times stated flat out that this was impossible and unsolvable and that we should just write a report about it.
I suspect those things have been trained by people who may not be quite as stubborn as I am.
But while the AI was ready to give up several times, it did keep adding debug code and analyzing it faithfully when I pushed. So credit where credit is due and I let the AI write the commit message above.
This is basically a one-liner fixing a bogus “round_up()” to a “round_down()”, but there were 24 patches adding more and more debug information to this, and 18 kernel boot to finally narrow it down to this. - Linus ]
Didn’t he state quite early on that LLMs were a good tool to spot bugs? Found this from 3 years ago: https://blog.mathieuacher.com/LinusTorvaldsLLM/
Regardless of how he got there, this is a pretty obnoxious and useless commit message…
Seriously. You don’t need to outline your whole journey to finding the bug, commentary on the tools you used, and how it all made you feel, just what the problem was and how it was fixed. Put the rest in a blog post.
Whatever’s the outcome, the anthropomorphization in that message stinks 🤮
Silly humans, antropomorphizing everything. Say, what species are you?
Correct. A being of power should never anthromorphise. It is a thing of flesh. I laugh at that because I am human and I never anthromorphise even the slightest part of a marble slab. I must be superior to most humans I guess. The forest people, though, those are legit just uninformed I think. I hate those creaks they make. I mean generate. I mean the wind . I mean no i mean the sound that nobody years because nobody is there and only humans and tree people are alive and the other animals are biological machines that we must try to ignore even if they lick us on the face wagging their tiny tail at us. For they are unworthy. I mean its tail. If it generates tail wags it is only an impression of aliveness and we must stand firm beside our tree people brothers and sisters on this.
Dude, I’m not an AI zealot, but I anthropomorphize my socks.
Trying to avoid anthropomorphising something with which you interact via natural language is incredibly laborious, and achieves nothing.
I like it.
Hackers use LLMs to spot odd exploits so it makes sense developers use them too. The usage of LLMs in development likely won‘t make Linux better or safer overall, because bugs and exploits wouldn‘t have been found by either side without it in the first place. It‘s like pandora‘s box where Linus is kind of forced to use it. If that makes sense.
Like an accelerating arms race.
He might consider himself forced to use it to find bugs. He wouldn’t be forced to use it to write the fix.
(Now you might say he wasn’t or isn’t, but even if that were true for the kernel at large this doesn’t seem to be: https://www.theregister.com/os-platforms/2026/08/10/linus-torvalds-says-ai-has-made-huge-linux-kernel-updates-the-new-normal/5285268 And this seems like mostly to be on Linus, for not stepping in here.)
Following the hacker arm race is a good point. But also the number of bugs is not a bottomless pit. If an LLM allows to find more, and hopefully the test and reviews are good enough to limit the new opportunities to introduce some from LLMs, it should reduce the overall number of bugs and make the software generally safer.
AI code bans may be impossible to enforce in the first place
I hate this attitude that if a rule or standard can’t be enforced perfectly, we shouldn’t strive for it. We can’t (reasonably) run an OS free of proprietary software, does that mean the free software movement should just give up?
“People are going to murder anyway, so why even make murder illegal?”
You are correct, you should make rules even understanding that they will be broken by some.
I hate this attitude that if a rule or standard can’t be enforced perfectly, we shouldn’t strive for it.
“Murder bans may be impossible to enforce in the first place”
I hate this attitude that if a rule or standard can’t be enforced perfectly, we shouldn’t strive for it.
This is the entire basis for the War on (some) Drugs. Guess what? Drugs won.
Power is nothing without enforceability, and purposely implementing rules or laws that can’t be enforceable is a form of malice and discrimination through selective enforcement.
This isn’t a case where something can’t be enforced perfectly, it is a case of something that can’t be enforced at all going into the future. Every tool that can reliably detect LLM code is at the same time the tool used for adversarial training, making the generated code look more and more human. We are already at the point where for plain english the false positives and false negatives go through the roof, making these tools very unreliable and when applied automatically a liability. Code is a lot more formalized, with a lot less personal variance (spelling, vocabulary and grammar are basically fixed - only the used logic and how it is implemented is variable), making detection harder by default than in natural languages.
If LLM code can’t be detected anymore by automated means - and that state of things is approaching fast - then any policy about allowing or restricting LLM code is not worth the paper you would use to print it out. But that’s not so much of a problem. The more important policy to set, that can also be enforced, is that everyone submitting code has to take personal responsibility regarding the quality of the submission. Delivering bad code - when not happening while training to become a better coder and looking for feedback - has to lead to consequences based on the seriousness of the case and if it’s a repeat offender. Anyone using an LLM to spit out bullshit LQ code will run into that kind of rule very fast.
I think one way is to make people realize that there is a concept in copyright called Threshold of Originality, meaning that because machine generated output in itself is not creative work, it’s not copyrightable. You can do literally anything with it, the licenses attached don’t matter. The GPL is not enforceable and no proprietary EULA is enforceable.
The entire point of GPL is to forcefully extract copyright out of the code in the first place.
No? The point of the GPL is to leverage such a copyright.
I think of it like age verification. Slipperly slope to a proprietary kernel-level “security compliance module” on all devices, aka the end of general computation.
AI code ban makes as much sense as code written in IDE ban. That’s why it is impossible to enforce. It makes no sense.
Me when I’m a logical fallacy
You’re getting a lot of down votes not for being wrong, but for being right.
AI psychosis swings both ways. There is definitely the c-suit managers who got marketed to and now fully embrace the insanity. Everything AI.
Then there are the people who think their feelings are valid logical reasons to start religious wars over. Apparently the existence AI insulted their ego and their “creativiae!” (in Cartman’s voice) by being better at writing and drawing than they are.
I would have more sympathy with the latter because they must be in incredible intellectual pain from disillusionment that all their poems and illustrations were in fact not art after all. But then the puritan loudmouths keep yapping on about what others can and cannot do. At this point I’m tempted to side with the Nerd Reich.
Aehm. I might be slightly affected as well.
I’m glad Linus is pragmatic about this, ethical questions aside LLM’s are a highly valuable tool in software development and they are likely to continue to improve. Any engineer fighting the use of LLM’s will ultimately find that their peers are achieving more in a shorter time then they are. Open weight LLM’s have also come a long way and are already good enough for daily use.
find that their peers are achieving more in a shorter time then they are
I’ve been reading this for 3 years and all I’ve seen so far is my peers making more work for me by going on wild goose chases and making me wade through slop. I do believe there are ways to use it responsibly and productively but that is definitely not what is happening on average.
I don’t understand why so many people decided to take “LLMs increase productivity” as an axiom instead of using their fucking eyes and looking at what is happening around them.
I remember people being angry at IDEs for making programming too easy and belittling code colourization as for children.
There will always be a crowd that wants everything to stay the way they know it.
It’s usually the illiterate that proclaim the sharpest opinions too. Linus is very steady and rational on this, he sees the operator as responsible and the real author. Which is I think how it should be. It reminds me of piracy debates in early internet. You have to be insane to think it is possible to prevent the use of a tool that you could just use and not say you used. It’s just futile and honestly just a counter psychosis of the same polarized issue. I just wanna tell both sides: Its not really a debate, because nobody is in charge or could ever enforce a decision either way. It’s also a form of privilegium overload to just think we can legislate or de-arm this knowledge and hunt the users down. It is an extreme position to think that it isn’t going to continue existing.
Some people still think that digital artists aren’t real artists.
Conservatives.
Any engineer fighting the use of LLM’s will ultimately find that their peers are achieving more in a shorter time then they are.
How do you measure that?
I am asking because I tried all the AI coding tools (copilot, cursor, claude code, etc.), even the paid versions, to find the utopia that doubles my productivity where lots of people are talking about. But everytime I find they take more time to generate code than me typing it, they rarely get things right the first time, they produce way more code than needed (which is much more to reason about) and all in all it takes longer to achieve something using “AI” instead of doing it on my own.
The worst results are the ones that seem to be correct, but can only be found incorrect on closer inspection. For instances, a part of my job is to write data collection applications for sensors; those sensors come with a datasheet (usually 100 to 200 pages) that explain how to “talk” to the sensor. Sounds perfect for an LLM - throw the datasheet in, get a small C application out. But either I get something that does not even compile, or something that seems to collect data, but configures the sensor completely wrong. Debugging that takes more time than just working through the datasheet myself and program it myself.
Yeah, my work is forcing us to use Claude for everything now. More than ever, I’m convinced that using AI is slower and worse than me just doing it by hand.
UNLESS, if I blindly accept the output, then Claude is faster. Sure, I can close the ticket in 15 minutes. Is it right? Idk. Is it good? Idk. Does it have bugs? Idk. Did it make shit up? Idk. Did anything even get compiled or tested? Idk. The ticket is closed though and that’s all that matters, apparently.
Well that’s because copilot and cursor are pure shit, and claude code is overpriced and over rated.
TBH I have been hopping between agent frameworks, but mostly I use locally run Qwen or Deepseek to draft individual function blocks or brainstorm in a notepad interface:
https://github.com/lmg-anon/mikupad
And manually insert them as I review the code… I like my stuff uncomplicated and unobstructed.
If I were to pick an API provider today, I’d buy one year of Xiaomi MiMo, because it is dirt cheap and good enough. GLM and Deepseek are great, but more expensive since they are popular now.
ethical questions aside
And his here is the problem. Oh yeah lets just waive off the biggest critisicm on and issue of AI.
Hitler was a pretty neat guy, ethnic cleansing aside.
And before you say this, no, i am not comparing you to hitler, I’m only using an over the top example to hopefully make my point more clear.
Wow, just speedrunning Godwin’s Law, aren’t we?
I guess I’m going to go back to work and talk to my five LLM sessions that are working on different tickets.
“Ai is literally Hitler”
-fediverse user, 2026
I think the better analogy is, the advances in rocketry made under hitler were pretty neat and just because Hitler was an evil guy who did evil shit doesn’t mean we should have thrown that away. By all means stop giving money and influence to the big AI companies, but the advances made can still be repurposed.
Will the Tesla be the next Beetle?
I’m only using an over the top example to hopefully make my point more clear.
If that’s the only example available that illustrates the point, then you have not spent enough time thinking through your idea.
Do better.
Yeah, ethical, environmental, legal and fuck the consumer memory questions aside.
That’s every corp and government in existence.
Well it’s about time those slackers caught up. And all they needed was a slop machine.
I think most misplaced critic comes from people not understanding or knowing the different types of AI. Some even hate all types of intelligence it seems.
Chess engines have been using AI, specifically, neural networks, for years before ChatGPT or other LLMs. Nobody complained.
Chess engines didn’t require data centers that guzzle water and spit out pollution like an 18th century factory.
Exactly right. That is the distinction I’m trying to draw.
Here we go again!
You conflate one type of AI with all the other ones. Call it LLM or generative AI.
I programmed AI 20 years ago, it still is AI but doesn’t use datacenters. I programmed deep neural networks ten years ago, it did not use datacenters either.
And that’s why no one is opposed to those.
You didn’t pirate the collective works of humanity to train on, while under a society that requires intellectual property rights to ensure those creators are compensated, in order to undercut them twice (once to train and once to replace them)
You didn’t pollute the sky, dry the soil, bulldoze people’s homes to make your chessbot.
You didn’t corner the market on components to brute force the solution to declare yourself computer god, where everyone must come to you for compute.
Gen AI seeks to replace humans and it’s not even sentient. I don’t give a single fuck how useful a fancy algorithm is, it’s not ok with me if it’s designed to destroy my life.
No but we did all that for Taylor swift and Walmart. Hopefully you don’t do anything with any major brands at all.
Well do the work and call it for what it is, not just “AI”.
Generative AI.
LLM
Is it truly that hard?
I mean you seem to want to, and thus we can have a discussion about it, which I favour instead of having people just being generally angry it seems.
The day when it doesn’t crush the environment, should we use gen AI to help us in our daily lives? Or should we not? Is there something good in it? Or isn’t it? I’d love seing constructive discussions but they just get globbed by people thinking with their gut it seems. Like if you don’t hate “AI” then you’re pro genocide.
Sorry for the rant!
They did, just on a smaller scale, and they used all games ever played by high-level players. LLMs and image generators were stealing data years before they became viable. And nobody gave a shit.
Let’s not pretend most people give a shit about ethics or the environment. Those were impacted long before AI became viable in any way. People hate on AI because it’s affecting everyday prices and employment. Ethics and environment are just convenient shiny arguments few actually stand behind.
If tomorrow billionaires turned around and said “actually, here’s UBI for everyone, provided by AI!”, most people would be cheering on intellectual theft and destruction of this earth, and call everyone still against it selfish and regressive.
Ethics and environmental destruction are “convenient shiny arguments”?? Go fuck yourself you delusional shill. Praying for a billionaire saviour while gagging on their boots, fuck you.
It’s all about scale, there aren’t billions of chess engine users.
Seems like an AI apologists article. Sounds a bit like “bans are hard so why even bother.”
No, the alternatives exist and people can choose them where able: https://codeberg.org/brib/slopfree-software-index Sure, not all things are available AI free, e.g. there’s no Linux kernel (NetBSD doesn’t run on as much hardware), but that doesn’t mean there’s no point in choosing no AI software where possible, if you care about it.
The enforceability part seems the most apologist to me. You never could really know if some contributor wasn’t copying leaked Windows XP code into your FOSS project. If you trust contributors that little, don’t let them contribute.
If the model is adequately FOSS, i.e. open weights, and can run on a single consumer GPU (or NPU), and the “author” (quotations because I’m personally undecided if one can claim code generated by an aforementioned model is theirs) understands it, then I really don’t see what the problem is…
Personally I have nothing against SLMs/LLMs as a technology, to me my grievances against ChatGPT or Claude are mostly about their environmental impacts and selling us back our own art, also keeping knowledge behind a for-profit black-box - if those aren’t appropriate for a specific model, then I say using that model is fair, and good for productivity.
I’d take a model whose training data is open source and legitimately obtained. The only ones I know about are Apertus and OLMo, and they aren’t really competitive.
Unfortunately not yet, no, true FOSS models are likely many years away, but I would argue that that follows typical FOSS lifecycles. Emerging technology is typically outperformed by proprietary endeavors, which creates an audience, and then that audience undertakes a FOSS implementation that initially underperforms, then just about competes, then eventually overtakes (i.e, GNU/Linux).
The problem with open weight models is that they are still hard to deploy. For DeepSeek for example you require NVIDIA HGX B200, these things are expensive.
Yes, unfortunately they are, and I do think efficiency is going to be a significant research front for open-weight models. The nature of this topic is highly speculative as our compute capabilities have only recently reached what is required to consider running generative AI models, what we have today are very crude first implementations of what I personally believe will become an everyday tool for developers, and more.
And we have seen this, there are models now capable of running on an individual’s hardware (and not particularly expensive hardware either) that can outperform what ChatGPT initially launched with.
This will also be quite the test for the Lemmy slopwarriors
Open Source and git mean this is basically a non issue. It’s just effort to manage it again.
People can fork a pre… 2020? version if they insist and if it’s enough people who want that, they can group up again and form a new community.
The problem AI slop coding is mostly the untested bullshit that causes problems. That’s not going to happen because I believe that a good bunch of stuff is thoroughly tested now, and also, this is maintainers doing it. If it doesn’t work, it’s their project, their career and their success on the line. And again, the worst possible outcome is that people have to roll back to an older version.
The code we use is not art the same way that gen-ai replaces “real artists”. I agree with opposing that tech in that area, and I agree with rejecting unchecked, untested, hallucinated nonsense in coding too, especially when it promises new features in corpo software that you are forced to use, and then have that fail on you.
But there will be a very normal selection process happening. If the process doesn’t work, the software will fail and lose users.
I fully support the people who absolutely reject it though, that means diversity of approaches and that’s good.
What was that random desktop image in the middle of the article for? It’s that the author’s? What kind of impression did he expect his elf babe wallpaper to convey?
