Comment on Linus Torvalds to critics of AI coding in Linux: "Fork it. Or just walk away."
pixxelkick@lemmy.world 2 months ago
Theres a forming deeply uninformed / misinformed loud group of people who desperately want to hate on AI being used for anything, even in scenarios where its actually perfectly good at its job.
Even when extremely seasoned professionals with waaaay more experience are stating “its a perfectly fine tool” these folks somehow cannot wrap their head around the possibility someone like fucking Linus Torvalds might know better than them.
Imagine having the utter balls to assert you know better than Torvalds about whether a technology is good or not 😐
Meanwhile those of us actually in the field have known for over a year now that the tools are pretty good now and very useful.
All major operating systems and software has been around before AI. Sure, AI might have some uses, even good ones but not enough to warrant all the negative effects it and all those datacenters running the models have.
Some people just don’t care about the negative effects and Torvalds is obviously one of them.
I hate AI as in, I hate Microsoft, Google, OpenAI, etc. for pushing it everywhere, being an oligopoly, and using it for mass surveillance. But AI as a tool is fantastic for some tasks
Every single generative tool is built on theft. No one has a training database that only contains code they were given legal permission to use for that purpose. Moreover, if these tools were trained on any GNU GPL code, then the tool and all of its output should also be GNU GPL:
if you distribute a derivative work or modification, you must provide the source code to those recipients under the same or equivalent license terms
en.wikipedia.org/wiki/GNU_General_Public_License
None of the companies developing these tools are obeying the terms of the licenses. The trained models are the product of theft.
You cannot be a moral person and approve of the use of these tools, they are diametrically opposed.
Hard disagree.
This is like saying anyone who has watched a Disney movie now owes Disney a check any time they illustrate anything, even entirely unrelated, because they’ve learned from their commercial products in the past. If they don’t pay Disney, then their creations are based on theft.
Fundamentally that’s a purely ethical decision about whether a machine doing something should be treated with the same understanding the same as a human would. But we don’t consider remembering something with a neurochemical storage the samd as taking a picture of something even if they’re basically the same.
That’s because taking a picture is an idetic copy. Machine learning is NOT copying, it’s learning - hence the name.
Digital reproduction is digital reproduction no matter how many extra steps are added to the reproduction process. It’s just an algorithm that sorts through a collection of stored data to find specific pieces of data which best fit the keywords supplied by the user, then regurgitates the results that are the best match based on correlation.
In spite of common meaning overlap and popular metaphors, the human brain is not a computer. Ask any neurologist.
Thank you for the opinion. If there’s a trial, and if you’re called as an expert witness, I’m sure they’ll consider your opinion, and have a fun debate on whether a human brain being similar to a computer is relevant in any way.
In the meantime, the technology exists, is useful and the results are no more theft than creating piñatas from newspaper clippings is theft from the newspaper.
Given that you’ve presumably looked at open source code in the past, and got more proficient (however marginally) as a result, does all your future code now belong to the GPL?
This comparison is invalid. Training a neural network algorithm is not equivalent to human learning. We are talking about data stored in machine learning models owned and controlled by multi-billion dollar corporations.
It has already been demonstrated multiple times that original training data can be reproduced completely from models, so yes, they are data storage systems. When they reproduce code which they have previously stored, even only in part, that is a derivative work. Adding extra steps to the transcribing process doesn’t make it any less a copy of the original.
Training a neural network algorithm is not equivalent to human learning
That’s a stretch, since we don’t know exactly how human learning works. Yes we have more than just that one mechanism at work obviously, but that’s not to say we’re not using the same or similar method as part of our learning process.
original training data can be reproduced completely from models
So can a few savants, more if you take into account those with special training, etc. The ability to do so is obviously latent to our brain.
When they reproduce code which they have previously stored, even only in part, that is a derivative work.
I don’t know about you, but I for one have not produced anything ‘original’ my entire coding career by that metric. I feel confident in saying the vast majority of programmers have not either.
Don’t mistake me, I’m well aware LLMs are not intelligent. But I disagree with the idea that their method of learning and their actions is inherently different from what the average person does.
Only when the date in question is so oblique as to the original source being the only place in which the data exists, which is neither common, nor useful in practice.
Copying is not theft.
Copying GNU GPL code without licensing your derivative code as GNU GPL and publishing it where it’s publicly accessible, and then using your derivative code to generate profit for a corporation, is definitely theft. Just because it’s open source doesn’t mean you can just do whatever you want with it, the original programmer still has rights over the code they wrote.
It literally isn’t.
Linus isn’t gonna let you hit little bro
Fuck AI
What a super useful addition to the convo.
Maybe think a little bit harder before hitting the reply button in the future, troll.
As opposed to your comment here?
Maybe ask ChatGPT to think harder next time.
To be fair, many have had their workplace enshittified by AI, or managers with AI psychosis.
Or fired over it.
…I think it’s understandable for people to be pissed, and take an absolutionist position, as the leads who see it as a tool seem to be few and far between.
Blaming AI for this, instead of the shitty people who made the decisions, is still very dumb.
I agree that aforementioned companies/managers/etc are shitty.
Thats not AIs fault lol, and isn’t a reflection of it.
Blaming AI as tge issue is just kneejerk mob mentality.
All true.
I’m just saying I can understand how people got in that mindset. I think the hate should redirected to the actual perpetrators: the Tech Bros.
I can agree with that.
Social media is a much bigger and dangerous threat to mankind than AI will ever be.
This is an extremely dangerous misconception.
I’m interested in reading your thoughts about it. Here are mine, in the best english I can muster.
I’m not saying AI is harmless. I’m saying social media has already demonstrated a more immediate and widespread ability to destabilize society. It has transformed how billions of people consume information, created echo chambers, rewarded putrage, accelerated misinformation, increase polarization, and eroded trust in shared facts. Again, AI mau eventuallt become more dangerous, but social media has already created the enviroment in which those dangers can spread. If you think as AI as a weapon, then social media is the distribution system, and we as a society have already given that distribution system to billions of people.
+1
Especially in regards to the nonsense that transformers LLMs will somehow lead to AGI. That is complete nonsense. The bubble is going to pop, and it’ll become a cheap utility.
Meanwhile, algorithmic optimization is basically the root of many world problems.
I’d say they’re in the same tier. It will take time to determine which is worse.
better than Torvalds
Have you forgotten about Bitkeeper?
Bravo! 💯
natecox@programming.dev 2 months ago including “the output is unacceptable bad”
Anyone with this view instantly goes in the “doesnt know wtf they are talking about, skill issue” bin.
“the tools are built in unethical and morally deficient ways”.
For coding, not really tbh. People are also deeply misinformed on this as well.
People are still operating on assumptions of how these tools work like its 2023 and not 2026.
The luddites haven’t been keeping up with the modern standards and probably couldnt even describe the nuance of when to choose a tool vs a skill for an optimization… let alone can speak as to what tools or skills to be using.
If you dont have opinions on how you’ve found stuff like caveman and headroom to work for you, I just straight up dont consider your opinion to be up to date enough to matter.
A big part of my job now is interacting with companies who have only just started to dip their toes into using these tools and so many developers are getting the sudden realization that keeping their head buried in the sand for the past 3 years has ruined their resume.
The difference between devs that at least tried to stay on top of things vs the luddites is palpable.
And on the total other end of the section you have fake devs cosplaying as real ones vis AI-everything but have 0 clue how to actually code.
Only the senior devs who both know how to code and how to use AI efficiently are aggressively outpacing everyone else.
Anyone with this view instantly goes in the "doesnt know wtf they are talking about, skill issue" bin.
Ah, I see painting everyone with one brush. Never mind the numerous studies stating otherwise. Oh the irony when you claim others are spreading misinformation.
“the tools are built in unethical and morally deficient ways”.
For coding, not really tbh. People are also deeply misinformed on this as well.
People are still operating on assumptions of how these tools work like its 2023 and not 2026.
So tell me, what has changed? Are we suddenly not gobbling Up the entire internet without permission to train AI models anymore? I must have missed something... (/s)
Honestly, personally, I tend to agree with Linus on this point, but holy fuck you have your hypocritical nose way up in the air.
Never mind the numerous studies stating otherwise
20 bucks says this person is basing their stance off that 1 study from a couple years ago where like a dozen developers with 0 experience with AI were tested using it and didnt show an increase in productivity.
Which is, in fact, precisely also what I was referring to when I said “People are still operating on assumptions of how these tools work like its 2023 and not 2026”
People use that study like its some big gotcha when in reality it just proves the simple fact that people need to be trained in how to use a tool in order for the tool to be useful.
So tell me, what has changed?
Dawg is indeed in the "doesnt know wtf they are talking about, skill issue” bin. Imagine unironically asking this question.
If you have to ask, you arent ready to participate in these convos and are just spreading a bunch of uninformed misinfo.
I dont want to talk to someone about this who hasnt put in the work to be informed.
Go get informed first, then come back in a few months and we can chat once you know the answer to that question.
natecox@programming.dev 2 months ago People truly are just exceptionally out of touch.
These convos constantly have people pulling out the tired old remarks.
Quoting 1 shitty study from years ago that is both heavily outdated and doesn’t support their argument the way they think it does, and their renarks indicate they didnt actually read the study
“But its theft” which is more of an ethical Grey discussion that Ill leave it up to courts to decide, and so far courts have not ruled AI usage for coding as illegal. Which means I do not give a shit. The entire concept is such obtuse ethical showboating. Only people grasping at straws try and pull this one out.
“But what about water usage” bitch we are software devs, data centers using water has been an issue for centuries do not pretend this is magically a new issue. If you have been deploying basically anything in the cloud, especially AWS, then picking AI as special makes you look like an incredibly uninformed idiot. Did you seriously think deploying entire docker containers constantly with integration tested pipelines on every PR wasnt also using water…?
“Its output is bad” skill issue. Literally skill issue. As in, “go learn how to use skill files”. In 2026 if your llm is still producing bad output thats 100% user error. Get good.
“But what about AIs deleting prod databases” yeah thats an even bigger user error. Why the hell did the LLM have access to be able to do that? Learn to properly use MCP tools and hiw to properly control what commands and operations you grant an LLM access to. Its not even that hard but people are just so lazy.
Every single goddamn thread people crawl out of the woodwork and hammer on repeating these same things. Its all they have.
If you actually wanna have a serious discussion maybe bring up more interesting dilemmas like Clean Room problems…
It’s not just Linus though. This is the exact position of many devs all over the world, ones who are still doing a good job and have clearly not fallen into AI psychosis.
Nobody really argues about that here though. The context of Linus statement is whether or not experienced kernel developers should be forbidden from using AI tools or not.
You may be surprised to learn that the maintainers of linux are not “untrained monkeys”
nagoya@quokk.au 2 months ago
The fact that some people don’t want to see AI being used to generate code does not make them dumb or misinformed. They may be informed, and simply reached a different conclusion than you.
Also, you have to consider that some of the most popular tools are controlled by a select few and that concentration can be used to control markets and users, reinforce biases, and shape what gets encouraged.
Finally, the use of AI can easily feed into a maximalist view of “produce more and faster” that has direct consequences like layoffs, increased cost of hardware, increased resource usage, increased environmental impact.
pixxelkick@lemmy.world 2 months ago
The majority of naysayers are touting deeply misinformed falsifities.
Its not a matter of guessing their intent, theyre quite outspoken on why they think its bad.
And the majority of them tout talking points that are years out of date at best.
Some valid points get brought up, but most of the points tend to be in the “these armchair devs have zero goddamn clue about how things actually are, they just are bandwagoning into an angry mob”
prole@lemmy.blahaj.zone 2 months ago
Did AI also tell you this was a word?