>The same process is pricing regular people out of the energy they need to power their air conditioner during those heat waves and the technology it enables is pushing them out of the labor market and into poverty.
I don't see evidence this is true, and the critical mass of data centers is making electricity cheaper. Economies of scale -> more energy = bigger electric plants = cheaper energy.
> Economies of scale -> more energy = bigger electric plants = cheaper energy.
Problem is; power plants are not something that can be spun up with a kubectl invocation.
Power plants need to be built. This process takes YEARS, and in some cases over a decade. And that's just the plants themselves. The grid also has to be built out to be able to handle the new load.
And while this potentially +10year process is ongoing, the current supply is gobbled up, and prices for consumers skyrocket.
Yes, eventuelly, market forces may solve the issue. By then however, it will already be too late for many areas, with residents being priced out of a basic utility.
Or worse, the markets may not solve the issue at all.
Because; all this building out of electrical gear and power plants, requires very specific resources and talent. Whta if the markets decide that these resources are much better allocated (at least from the PoV of company shareholders), in building power generation capacity exclusively for datacenters, and the consumer market is simply ignored?
> Economies of scale -> more energy = bigger electric plants = cheaper energy.
This isn't true. Certainly not globally. Because electricity in 2026 is converting finite and scares resources. Plants need fuel to burn, water to cool, material to build, land to sit on, etc. Yes, even Nuclear plants¹ need increasingly more scarce resources like "coolant".
What has been bringing electricity prices down is renewables: insane amounts of solar and wind being installed hourly. But even with that, we are still lagging, globally: energy prices are -on average- rising, projects halted or stalled because of lack of electrical capacity and ever increasing amounts of CO2 pumped into the atmosphere. Not less. Not cheaper. Not more supply.
It is true globally as well, but there are certainty anecdotes where confounding factors makes it not true like what you've highlighted with specific France Nuclear reactors.
We should be cognizant of the environmental impact of producing more energy, but it still remains true that making more energy means cheaper energy across the resources we have today.
From the title, I was hoping this was going to be a story about the major AI distros agreeing on an API and basic AI capabilities to be included in Linux just like Microsoft and Apple have done with their operating systems.
Instead, it's:
> how many tons of CO₂ added to the atmosphere or liters of fresh water supplies disrupted is a tolerable price for a better code review?
The article isn't trying to find a particular example. It's talking about the overall stance of Linux on AI use. It argues that maintainers should consider political/moral/ethical arguments around climate change, effect on water supplies, energy prices, etc.
AI is just a tech like any other. Can be powered by CO2-neutral electricity like anything else. Otherwise, we'll have to protest against aluminum cans, airplanes, golf courts, lemon trees, and whatever else that uses water or electricity.
That's a false equivalence. AI is not the same as aluminum cans.
All technologies have political, social, ethical, and economic aspects, and not all technologies are the same. AI is not, in reality, powered by CO2-neutral electricity. As the article mentions, AI is the fastest-growing energy user in the world, so it probably merits a deeper look than aluminum cans.
Incredibly useful for some definition of incredibly useful. What I see:
- I now regularly receive AI-written spam emails (marketing pitches, "vulnerability reports").
- I was receiving AI-written garbage quality pull requests. It got so annoying I closed pull requests on my project's repo.
- When I try to contact some company's customer support these days there's a good chance I will be fed slop by an agent posing as a customer support person, this is absolutely infuriating.
- When I search for information on the internet, for certain queries, 90% of the results is slop.
- AI crawlers are DDoSing websites.
- The already mentioned things – hardware getting more expensive, climate impact, unethically sourced training data.
True in fact. Aluminum manufacturing is responsible for roughly 2% of global greenhouse gas emissions, and if you add in metal manufacturing as a whole it comes up to 11%. Refining ore is very energy-intensive.
Aluminum cans had a distinct use. AI just invades a perfectly working industry that made Torvalds filthy rich and adds nothing- And he defends it because he is corporate now.
Why not fix all the issues with private jets and yachts first? Why don't they use regular public transportation as we all do? Instead, let's bash poor Clanker, who reviews kernel code.
The thing I'd worry about is it slowly handing over control of Linux's direction to the tech companies who control and regulate the AI being used to help develop it. The complexity is only growing. These companies are themselves extremely political and decide what AI is and isn't legal when Linux demands using the most capable models for security reasons. So it would be good at least if the project actually cared and had it's own principles besides ignoring the politics.
But the code being GPL means those companies don't have control over it. They can restrict the use of AI but the codes already generated, they would just go back to writing by hand. Its a non-issue. The same objection came up when Microsoft started contributing to the kernel.
And despite the author’s logical leaps; no it's not impossible to fork the kernel and go another route. The fact it doesn't happen is because its stewardship so far has been exceptional.
It's not a political issue. It's a religious one with the anti AI people raising pitchforks.
Temperatures during heat waves are exceeding 50°C in India, causing tens thousands of deaths. The AI built-out is by far the fastest growing energy consumer in the world, and they’re being built with fossil fuels, or drawing green energy demand away from replacing the fossil fuels depended on by other industries.
I hate how this ostentatious caring for Indians is used to push personal political goals. Want to help with global warming? Reduce your consumption. Stop driving, stop eating non-local food, stop buying things made off-shores or with supply chains that require global shipping. Move to a much poorer country (or just a poorer neighbourhood of your country) and live the way people there live.
It seems like for much of Western world, Indians aren't really humans but a giant undifferentiated mass of temporarily animated bodies that are fit only as a canvas for one's own political paintwork.
The author seems to be making a mountain out of a molehill, to be honest. Linux has never been a "political project" like the author asserts. GNU is closer to that, but even GNU has a purely technical component (the software itself).
If everything is political, as DDV likes to rant every now and then, then Linus' politics are those related to how software is to be developed and distributed. I don't see why a software project _should_ care about global warming, minorities, state policies and other issues outside of its scope. It could, and it can be a good thing to do, but it doesn't need to.
It feels wrong to me, though, to see AI being used in kernel development, as I see the trouble it causes on regular projects.
> Looking at it from another angle: how many tons of CO₂ added to the atmosphere or liters of fresh water supplies disrupted is a tolerable price for a better code review?
He should, but you'll never get these free software wingnuts to see logic. They back-rationalize everything from ideologically drawn conclusions.
And they're driven by fear--fear of being replaced. This is one reason why open source projects are limiting use of LLMs. LLMs finally make software free! Or at least, much cheaper.
In reality, these reactionary anti-AI policies mean that users are private forking OSS projects, improving them with AI, and not contributing the changes back, making the original OSS project even more irrelevant over time.
This is an important conversation to have and we should pretty clearly about what the issues are.
I didn't see any objection to LLM assisted development in the abstract. The objection is to relying on proprietary AI providers in general, and Gemini in particular, via Shashiko. The problem is that kernel contributions could become gated by hyperscalers who can draw on compute power normal people can't access.
Devault mentions data centers' carbon emissions and water usage. I find both of these objections to be distractions. It's not that carbon emissions don't matter. Keeping greenhouse gasses in check is incredibly important. But there are far more productive ways to cut carbon emissions than to block data centers. Heavy industry consumes far more electricity than data centers, for less economic value. This is even truer for water use. If carbon emissions is a major concern (as it should be), there are better targets to go after than data centers.
A more compelling argument is the danger that the Linux kernel becomes something that's effectively controlled by proprietary AI providers: Google, Anthropic, OpenAI, Microsoft, etc. The photo of tech company leaders at Trump's inauguration underscores their lack of any moral compass.
But if AI-driven kernel development is actually useful and productive, that's going to be hard to turn down. If you believe that LLM use is technically counter-productive, then this whole AI thing is a bubble and you can just wait for it to pop. If you believe that LLMs can be technically productive if used correctly, and the drawbacks are entirely social, then it means giving up something technically useful. That is a real cost.
I think Linus is right here: it would be hard for regular kernel developers to coordinate a fork of the project. It might not be that hard for the hyperscalers to simply fork Linux. They would still release the source code under the GPL, but nobody would be able to keep up with it. Ultimately, this could doom the human-written version of the kernel to irrelevance as most people switch to the faster/stronger/harder LLM-derived fork.
Maybe a reasonable medium would be to insist on using an open model for code review instead of depending on Gemini or any proprietary model.
I do not see how AI could be legally used on a GPL project.
But if you want to use a non-AI OS, NetBSD banned the use of AI. Maybe some of the talent will move over there. They can sure use the help.
As for AI in Linux, I am on the fence about that, but since Linux is pretty much funded/owned by large corporations, AI being forced on Linux is no surprise at all.
The Office concludes that, given current generally available technology, prompts alone do not provide sufficient human control to make users of an AI system the authors of the output.
I'm not a lawyer either but if/when something like this goes to court the conclusion of the US Copyright Office is probably going to play a part.
What I was able to skim is a bit more complex like that. Page 42 of Part 2 says using it "as a tool" produces copyrightable works:
> As described above, in many circumstances these outputs will be copyrightable in whole or in part—where AI is used as a tool, and where a human has been able to determine the expressive elements they contain. Prompts alone, however, at this stage are unlikely to satisfy those requirements.
Then, since this stuff is changing a lot, they hedge with:
> The Office continues to monitor technological and legal developments to evaluate any need for a different approach.
I'm not a lawyer, but based on everything I've read on the topic I think there's a solid argument that there are legitimate questions about the use of copyrighted material in training data, and whether or not the output of an LLM should be considered a derived work. As I googled it, it seems like courts have mostly called it OK but maybe this matter isn't 100% settled or converged upon a single answer.
Linus' statement that Linux is not political is strange. It started as a resistance project and used GPL for ideological reasons.
Linus himself was made to sign the corporation-enforced CoC and "retreat" for "counseling".
He came back a changed man. Since then I've only see him a) use the CoC as a weapon for the old boys and b) parrot any current mainstream or corporate line.
That includes COVID vaccines, overt anti-Russian statements (when not warranted) and now AI. Would he still give Ndivia the finger? I doubt it.
These Gen-X millionaires cling to their positions and I'm sure the Linux Foundation has a well oiled HR department. Ted Tso also already interfered pro AI in the Debian discussion, where the pro-AI resolution is also championed by Canonical shills.
tl;dr Kernel devs should prompt less because it uses up water or something.
Repeat after me: you can't put the cat back in the bag. Pandora's box is open. Sucks, but you have to learn to live with the reality that most developers will use AI to write code from now on.
How do I know if comments like this come from a real person or a bot?
Your suggestion that "anything popular cannot be morally wrong" is proposterous. It feels like you're trying to bait people into arguing with you in a way I associate with bot spam.
But on the off change you're a real person - I will politely say - just because something is useful and popular doesn't meant it doesn't have negative externalities and doesn't mean people shouldn't fight to make it better.
Improving emissions standards is important even though lots of people drive cars. The same can absolutely be true of AI.
> Your suggestion that "anything popular cannot be morally wrong" is proposterous.
It’s not about being popular, it’s about facing a new economic reality. Once machines were invented there was no going back to human labor, no matter how bad it was for the displaced workers or the environment.
Technology has always been a relentless economic force. And this is largely because despite the myriad externalities, it’s blindingly obvious that we as a species are, on average, much better off with technology than without.
So, instead of getting upset at the people taking a joyride on the gas guzzling Lamborghini or generating shitty slop memes on TikTok, you're upset the ambulance drivers are racing to help the world out.
> How do I know if comments like this come from a real person or a bot?
Presumably, I'm a bot because you disagree with me.
> Your suggestion that "anything popular cannot be morally wrong" is proposterous.
> just because something is useful and popular doesn't meant it doesn't have negative externalities and doesn't mean people shouldn't fight to make it better.
PREposterous. And I said nothing of the sort. It feels like you're hallucinating things nobody said in a way I strongly associate with bot spam.
> Improving emissions standards is important even though lots of people drive cars. The same can absolutely be true of AI.
The article did not at any point suggest that we need to improve and optimize AI. There is no magic wand that Linus can wave to measure how much water the incoming AI-assisted contributions are using. The guy should just go protest at a data center construction site or something; the article is pointless and completely fails to diagnose the actual problems with AI or propose viable solutions.
This is a really disingenuous way to characterize the article (CO2 and global warming were much more of a focus than water, so your "or something" is doing a lot of work).
Furthermore your comment ignores key arguments in the article, particularly the argument that contributors should not be forced to participate in AI-based ceremonies such as LLM code review that they may object to on personal ethical grounds.
> This is a really disingenuous way to characterize the article (CO2 and global warming were much more of a focus than water, so your "or something" is doing a lot of work).
Remind me please, how much exactly do data centers contribute to greenhouse emissions? And how much something like, I don't know, meat production contribute?
If my Googling is correct, apparently a single ChatGPT query produces ~0.004 kg CO2e/kg. Beef production contributes ~60 kg CO2e/kg. So you'd need to make, err... over 15k ChatGPT queries to produce as much emissions as eating a single kilogram of beef.
I've seen anti-AI protests. Where are all the anti-meat protests? I don't mean to engage in whataboutism, I'm just saying -- if your goal is to reduce CO2 in the atmosphere then there are bigger fish to fry, instead of just virtue signaling because there's a new trendy bogeyman in town.
What's with this whataboutism? What if I told you I was vegan? What if I told you I only drive to work or to medical appointments? Would you pull another excuse out to justify your emissions?
I've seen anti-AI protests. Where are all the anti-meat protests?
Kernel developers just aren’t emitting that much CO2 with their proompting. It’s okay that you’re pissed off, but it would be more productive if you were pissed off by real world issues.
And it's not Linux' task to close that, even if it could, which it cannot. How is that disingenuous?
I agree that we should keep discussing the merits and dangers of AI-aided software development. Broad, so indeed also CO2 etc. But that discussion is not something that Linux should push forward. It's a discussion that should take place outside the context of specific projects and in the context of the AI tools and providers, IMO.
It's disingenuous because it's attacking something that the article didn't say. It didn't say "kernel devs should prompt less because it uses up water."
The real argument was that AI has many externalities that should be considered (climate change, water use, energy prices, putting people out of jobs) and that Linux, as a relevant and influential political actor, should engage with these topics. It's not even giving a final determination about what Linux should do, it's just disagreeing with Torvalds' total rejection of ethical arguments about AI.
Even if you believe that the technical merits outweigh the other costs, it seems at least reasonable to engage on those grounds. But Torvalds isn't doing that.
Also, simply stating that the "cat is out of the bag" and "Pandora's box is open" isn't really an argument. These are thought-terminating cliches; not logical arguments
> These are thought-terminating cliches; not logical arguments
Consider sometimes terminating some thoughts and replacing them with actually useful ones. It’s good for you and your goals regardless of what they are.
People burning thought-hours on pretending that we can make AI go away are basically doing nothing except emitting CO2 and consuming water. If you want change, start with yourself!
Anyway, so far zero actionable suggestions for Torvalds in the article or in this thread. Great job so far.
Contributors may object to the Hitler AI on their personal ethical grounds, but it's not going to make their buggy and vulnerable code any less so. Because our societies critically depend on projects like Linux, it follows that someone willing to pick up the pieces and do the work will then take their place.
There are many arguments one could make in favor of keeping AI around and using it, but your "argument" here is just about the lousiest and most naive argument one could possibly make, and shows a complete lack of empathy and understanding from your end.
Torvalds' statement was incredibly disingenuous and rather sad. Every single commit is making a legal and political claim. Each commit is saying "this code is my original copyrighted work, and I am licensing it to the world per the requirements of the GNU GPLv2" (or "this code is someone else's work which was licensed under a compatible license, and I am integrating it into this GNU GPLv2 project").
That's not social warrior, that's simple facts. We don't need to try and guess motives or reasoning for people's contributions - that doesn't matter. What we have is an explicit declaration by virtue of the license, that the code being provided can legally be included in the GNU GPLv2 licensed "Linux" kernel. But with LLM generated code, this legal claim is dubious at best. For trivial (!= simple) code, one can probably argue that the LLM code will not be violating anyone's copyright, however for anything remotely novel, unless you are 100% certain that all of the training data was GPLv2 or compatible, it is not legally safe to use - and aside from that, even if all the training data was compatible with the license, the LLM will virtually always fail to properly attribute the source of its output, which is a requirement of - as far as I can tell - every license compatible with the GNU GPLv2. I cannot share GPL'd code and misrepresent it as my original work, I still have to make clear that it was someone else's creation.
LLM generated code is a ticking time bomb, and is going to poison projects.
Yes, it is possible to do this already without LLMs, but not by accident. A developer would have to make a conscious decision to violate someone's license in the past. Now it can happen without anyone being the wiser.
And as someone who very much uses GNU GPL licensed software because of its license, this worries me for the future of software. On technical merits, it doesn't matter, but if you care at all about the artful expression of software, as well as honesty about who created what, this should concern you.
> LLM generated code is a ticking time bomb, and is going to poison projects.
Given the recent Anthropic ruling and OAI settlements, and the general stance of courts to look the other way when it comes to LLM use, I think an exemption for LLM-based software will be handwaved into existence (to my regret.)
That certainly seems to be the way the courts are going, which is also quite troubling, considering a plain understanding of software licenses and how LLMs work should clearly demonstrate the issues I outlined above.
I don't want to accuse the courts of intentionally siding with big corporations (I have no evidence to support such a claim anyway), but the only other option I see is that the courts are uninformed on either (A) the technology at play, or (B) the requirements of many software licenses.
Though, this should also apply to works licensed under variants of the Creative Commons (with Attribution), since once again, you can end up using parts of a work that requires attribution without even knowing that it's someone else's work (or at least without knowing its license requirements). There might be a better argument for generative AI images being transformative (though I think still a weak argument since you could almost certainly produce a 1:1 reproduction of a piece of training data, much like LLMs have already been pushed to doing with snippets from books).
>The same process is pricing regular people out of the energy they need to power their air conditioner during those heat waves and the technology it enables is pushing them out of the labor market and into poverty.
I don't see evidence this is true, and the critical mass of data centers is making electricity cheaper. Economies of scale -> more energy = bigger electric plants = cheaper energy.
https://netchoice.org/new-study-confirms-what-weve-been-sayi...
> Economies of scale -> more energy = bigger electric plants = cheaper energy.
Problem is; power plants are not something that can be spun up with a kubectl invocation.
Power plants need to be built. This process takes YEARS, and in some cases over a decade. And that's just the plants themselves. The grid also has to be built out to be able to handle the new load.
And while this potentially +10year process is ongoing, the current supply is gobbled up, and prices for consumers skyrocket.
Yes, eventuelly, market forces may solve the issue. By then however, it will already be too late for many areas, with residents being priced out of a basic utility.
Or worse, the markets may not solve the issue at all.
Because; all this building out of electrical gear and power plants, requires very specific resources and talent. Whta if the markets decide that these resources are much better allocated (at least from the PoV of company shareholders), in building power generation capacity exclusively for datacenters, and the consumer market is simply ignored?
The largest build-out, which is in renewables, features plants that can go up very quickly, at least. The grid upgrades are the bigger bottleneck.
> Economies of scale -> more energy = bigger electric plants = cheaper energy.
This isn't true. Certainly not globally. Because electricity in 2026 is converting finite and scares resources. Plants need fuel to burn, water to cool, material to build, land to sit on, etc. Yes, even Nuclear plants¹ need increasingly more scarce resources like "coolant".
What has been bringing electricity prices down is renewables: insane amounts of solar and wind being installed hourly. But even with that, we are still lagging, globally: energy prices are -on average- rising, projects halted or stalled because of lack of electrical capacity and ever increasing amounts of CO2 pumped into the atmosphere. Not less. Not cheaper. Not more supply.
¹ https://www.france24.com/en/france/20260712-france-temporari...
It is true globally as well, but there are certainty anecdotes where confounding factors makes it not true like what you've highlighted with specific France Nuclear reactors.
We should be cognizant of the environmental impact of producing more energy, but it still remains true that making more energy means cheaper energy across the resources we have today.
Yeah, I wouldn't have that paper peer reviewed either. Garbage methodology.
netchoice is a lobbying organization for the hyperscalers, you're straight up sharing a propaganda piece
Which metric do you disagree with presented on the site?
From the title, I was hoping this was going to be a story about the major AI distros agreeing on an API and basic AI capabilities to be included in Linux just like Microsoft and Apple have done with their operating systems.
Instead, it's:
> how many tons of CO₂ added to the atmosphere or liters of fresh water supplies disrupted is a tolerable price for a better code review?
The answer isn't zero, is it?
And also it can be zero. Nuclear power is CO2 neutral. So is solar.
I read through this and must have missed it, but I don't see a single example of AI use in the Linux kernel development posing a problem.
The article isn't trying to find a particular example. It's talking about the overall stance of Linux on AI use. It argues that maintainers should consider political/moral/ethical arguments around climate change, effect on water supplies, energy prices, etc.
AI is just a tech like any other. Can be powered by CO2-neutral electricity like anything else. Otherwise, we'll have to protest against aluminum cans, airplanes, golf courts, lemon trees, and whatever else that uses water or electricity.
Plenty of people campaign against things like airplanes or golf courses. They might even be the same folks protesting AI usage! Imagine that!
That's a false equivalence. AI is not the same as aluminum cans.
All technologies have political, social, ethical, and economic aspects, and not all technologies are the same. AI is not, in reality, powered by CO2-neutral electricity. As the article mentions, AI is the fastest-growing energy user in the world, so it probably merits a deeper look than aluminum cans.
AI is also an incredibly useful tool that could automate many industries, so it deserves some leeway because of the benefits it provides.
And while it is 'fastest-growing', it is small as an absolute share - around 0.2% of total world CO2 emissions.
So it can both outprice people as well as take their jobs? Well, leeway away.
The only reason we have the standard of living we do today is because of automation. I welcome more of it.
Incredibly useful for some definition of incredibly useful. What I see:
- I now regularly receive AI-written spam emails (marketing pitches, "vulnerability reports").
- I was receiving AI-written garbage quality pull requests. It got so annoying I closed pull requests on my project's repo.
- When I try to contact some company's customer support these days there's a good chance I will be fed slop by an agent posing as a customer support person, this is absolutely infuriating.
- When I search for information on the internet, for certain queries, 90% of the results is slop.
- AI crawlers are DDoSing websites.
- The already mentioned things – hardware getting more expensive, climate impact, unethically sourced training data.
Aluminum production is one of the largest electricity consumers out there, much larger than AI.
True in fact. Aluminum manufacturing is responsible for roughly 2% of global greenhouse gas emissions, and if you add in metal manufacturing as a whole it comes up to 11%. Refining ore is very energy-intensive.
Aluminum cans had a distinct use. AI just invades a perfectly working industry that made Torvalds filthy rich and adds nothing- And he defends it because he is corporate now.
It is more like leaded fuel or DDT.
Adds nothing? In what world are you living?
Why not fix all the issues with private jets and yachts first? Why don't they use regular public transportation as we all do? Instead, let's bash poor Clanker, who reviews kernel code.
The thing I'd worry about is it slowly handing over control of Linux's direction to the tech companies who control and regulate the AI being used to help develop it. The complexity is only growing. These companies are themselves extremely political and decide what AI is and isn't legal when Linux demands using the most capable models for security reasons. So it would be good at least if the project actually cared and had it's own principles besides ignoring the politics.
But the code being GPL means those companies don't have control over it. They can restrict the use of AI but the codes already generated, they would just go back to writing by hand. Its a non-issue. The same objection came up when Microsoft started contributing to the kernel.
And despite the author’s logical leaps; no it's not impossible to fork the kernel and go another route. The fact it doesn't happen is because its stewardship so far has been exceptional.
It's not a political issue. It's a religious one with the anti AI people raising pitchforks.
Tens of thousands of people dying as a direct result of global warming! Not a problem! /s
No ones dead because Linux devs prompted a chat bot
Exactly. When I change my engine oil, I just dump it into the street for the same reason - nobody is dying from my actions directly.
No one is dead because I launched a nuclear missile.
That doesn't mean I _should_ be given access to a nuclear missile.
Temperatures during heat waves are exceeding 50°C in India, causing tens thousands of deaths. The AI built-out is by far the fastest growing energy consumer in the world, and they’re being built with fossil fuels, or drawing green energy demand away from replacing the fossil fuels depended on by other industries.
https://www.worldometers.info/co2-emissions/co2-emissions-by...
Share of global CO2:
China 33%
US 12%
India 8%
Russia 5%
Japan 3%
History didn't start in the last 12 months.
What's your point?
I hate how this ostentatious caring for Indians is used to push personal political goals. Want to help with global warming? Reduce your consumption. Stop driving, stop eating non-local food, stop buying things made off-shores or with supply chains that require global shipping. Move to a much poorer country (or just a poorer neighbourhood of your country) and live the way people there live.
It seems like for much of Western world, Indians aren't really humans but a giant undifferentiated mass of temporarily animated bodies that are fit only as a canvas for one's own political paintwork.
What makes you think I don't do all that stuff?
This is a “everything is political” rant but now with 50% more AI.
like we could expect something other than that from DDV
The author seems to be making a mountain out of a molehill, to be honest. Linux has never been a "political project" like the author asserts. GNU is closer to that, but even GNU has a purely technical component (the software itself).
If everything is political, as DDV likes to rant every now and then, then Linus' politics are those related to how software is to be developed and distributed. I don't see why a software project _should_ care about global warming, minorities, state policies and other issues outside of its scope. It could, and it can be a good thing to do, but it doesn't need to.
It feels wrong to me, though, to see AI being used in kernel development, as I see the trouble it causes on regular projects.
I propose we get rid of tiktok videos and let the kernel devs use AI. That should balance.
> Looking at it from another angle: how many tons of CO₂ added to the atmosphere or liters of fresh water supplies disrupted is a tolerable price for a better code review?
Fewer than if a human had done it without AI: https://news.ycombinator.com/item?id=48941143
I guess that means he should advocate for more use of AI for Linux dev! ;-)
He should, but you'll never get these free software wingnuts to see logic. They back-rationalize everything from ideologically drawn conclusions.
And they're driven by fear--fear of being replaced. This is one reason why open source projects are limiting use of LLMs. LLMs finally make software free! Or at least, much cheaper.
In reality, these reactionary anti-AI policies mean that users are private forking OSS projects, improving them with AI, and not contributing the changes back, making the original OSS project even more irrelevant over time.
This is an important conversation to have and we should pretty clearly about what the issues are.
I didn't see any objection to LLM assisted development in the abstract. The objection is to relying on proprietary AI providers in general, and Gemini in particular, via Shashiko. The problem is that kernel contributions could become gated by hyperscalers who can draw on compute power normal people can't access.
Devault mentions data centers' carbon emissions and water usage. I find both of these objections to be distractions. It's not that carbon emissions don't matter. Keeping greenhouse gasses in check is incredibly important. But there are far more productive ways to cut carbon emissions than to block data centers. Heavy industry consumes far more electricity than data centers, for less economic value. This is even truer for water use. If carbon emissions is a major concern (as it should be), there are better targets to go after than data centers.
A more compelling argument is the danger that the Linux kernel becomes something that's effectively controlled by proprietary AI providers: Google, Anthropic, OpenAI, Microsoft, etc. The photo of tech company leaders at Trump's inauguration underscores their lack of any moral compass.
But if AI-driven kernel development is actually useful and productive, that's going to be hard to turn down. If you believe that LLM use is technically counter-productive, then this whole AI thing is a bubble and you can just wait for it to pop. If you believe that LLMs can be technically productive if used correctly, and the drawbacks are entirely social, then it means giving up something technically useful. That is a real cost.
I think Linus is right here: it would be hard for regular kernel developers to coordinate a fork of the project. It might not be that hard for the hyperscalers to simply fork Linux. They would still release the source code under the GPL, but nobody would be able to keep up with it. Ultimately, this could doom the human-written version of the kernel to irrelevance as most people switch to the faster/stronger/harder LLM-derived fork.
Maybe a reasonable medium would be to insist on using an open model for code review instead of depending on Gemini or any proprietary model.
I do not see how AI could be legally used on a GPL project.
But if you want to use a non-AI OS, NetBSD banned the use of AI. Maybe some of the talent will move over there. They can sure use the help.
As for AI in Linux, I am on the fence about that, but since Linux is pretty much funded/owned by large corporations, AI being forced on Linux is no surprise at all.
Are you a lawyer or an interlectual-property legal expert?
Because if so, please, please elaborate on
> I do not see how AI could be legally used on a GPL project.
Too many people "believe" or "are convinced" about matters with AI and Open Source without any backup or sources.
It's because, at least in the US, AI generated code cannot be protected by copyright.
<https://www.copyright.gov/ai/>
<https://www.copyright.gov/ai/Copyright-and-Artificial-Intell...>
I'm not a lawyer either but if/when something like this goes to court the conclusion of the US Copyright Office is probably going to play a part.What I was able to skim is a bit more complex like that. Page 42 of Part 2 says using it "as a tool" produces copyrightable works:
> As described above, in many circumstances these outputs will be copyrightable in whole or in part—where AI is used as a tool, and where a human has been able to determine the expressive elements they contain. Prompts alone, however, at this stage are unlikely to satisfy those requirements.
Then, since this stuff is changing a lot, they hedge with:
> The Office continues to monitor technological and legal developments to evaluate any need for a different approach.
I'm not a lawyer, but based on everything I've read on the topic I think there's a solid argument that there are legitimate questions about the use of copyrighted material in training data, and whether or not the output of an LLM should be considered a derived work. As I googled it, it seems like courts have mostly called it OK but maybe this matter isn't 100% settled or converged upon a single answer.
Because you can't slap a copyright on AI output.
Linus' statement that Linux is not political is strange. It started as a resistance project and used GPL for ideological reasons.
Linus himself was made to sign the corporation-enforced CoC and "retreat" for "counseling".
He came back a changed man. Since then I've only see him a) use the CoC as a weapon for the old boys and b) parrot any current mainstream or corporate line.
That includes COVID vaccines, overt anti-Russian statements (when not warranted) and now AI. Would he still give Ndivia the finger? I doubt it.
These Gen-X millionaires cling to their positions and I'm sure the Linux Foundation has a well oiled HR department. Ted Tso also already interfered pro AI in the Debian discussion, where the pro-AI resolution is also championed by Canonical shills.
You can't fork, but you can use BSD.
Drew, there will be new jobs.
Also, the open models are a great alternative to the datacenter ones.
I imagine AI will go towards the local models more when the bubble pops.
tl;dr Kernel devs should prompt less because it uses up water or something.
Repeat after me: you can't put the cat back in the bag. Pandora's box is open. Sucks, but you have to learn to live with the reality that most developers will use AI to write code from now on.
How do I know if comments like this come from a real person or a bot?
Your suggestion that "anything popular cannot be morally wrong" is proposterous. It feels like you're trying to bait people into arguing with you in a way I associate with bot spam.
But on the off change you're a real person - I will politely say - just because something is useful and popular doesn't meant it doesn't have negative externalities and doesn't mean people shouldn't fight to make it better.
Improving emissions standards is important even though lots of people drive cars. The same can absolutely be true of AI.
> Your suggestion that "anything popular cannot be morally wrong" is proposterous.
It’s not about being popular, it’s about facing a new economic reality. Once machines were invented there was no going back to human labor, no matter how bad it was for the displaced workers or the environment.
Technology has always been a relentless economic force. And this is largely because despite the myriad externalities, it’s blindingly obvious that we as a species are, on average, much better off with technology than without.
So, instead of getting upset at the people taking a joyride on the gas guzzling Lamborghini or generating shitty slop memes on TikTok, you're upset the ambulance drivers are racing to help the world out.
> How do I know if comments like this come from a real person or a bot?
Presumably, I'm a bot because you disagree with me.
> Your suggestion that "anything popular cannot be morally wrong" is proposterous. > just because something is useful and popular doesn't meant it doesn't have negative externalities and doesn't mean people shouldn't fight to make it better.
PREposterous. And I said nothing of the sort. It feels like you're hallucinating things nobody said in a way I strongly associate with bot spam.
> Improving emissions standards is important even though lots of people drive cars. The same can absolutely be true of AI.
The article did not at any point suggest that we need to improve and optimize AI. There is no magic wand that Linus can wave to measure how much water the incoming AI-assisted contributions are using. The guy should just go protest at a data center construction site or something; the article is pointless and completely fails to diagnose the actual problems with AI or propose viable solutions.
if you suspect someone is a bot, email the mods.
if you're just saying it as an ad hominem, that is pretty lame, and you can make a stronger argument without it.
This is a really disingenuous way to characterize the article (CO2 and global warming were much more of a focus than water, so your "or something" is doing a lot of work).
Furthermore your comment ignores key arguments in the article, particularly the argument that contributors should not be forced to participate in AI-based ceremonies such as LLM code review that they may object to on personal ethical grounds.
> This is a really disingenuous way to characterize the article (CO2 and global warming were much more of a focus than water, so your "or something" is doing a lot of work).
Remind me please, how much exactly do data centers contribute to greenhouse emissions? And how much something like, I don't know, meat production contribute?
If my Googling is correct, apparently a single ChatGPT query produces ~0.004 kg CO2e/kg. Beef production contributes ~60 kg CO2e/kg. So you'd need to make, err... over 15k ChatGPT queries to produce as much emissions as eating a single kilogram of beef.
I've seen anti-AI protests. Where are all the anti-meat protests? I don't mean to engage in whataboutism, I'm just saying -- if your goal is to reduce CO2 in the atmosphere then there are bigger fish to fry, instead of just virtue signaling because there's a new trendy bogeyman in town.
What's with this whataboutism? What if I told you I was vegan? What if I told you I only drive to work or to medical appointments? Would you pull another excuse out to justify your emissions?
I've seen anti-AI protests. Where are all the anti-meat protests?
Fucking everywhere bro.
Kernel developers just aren’t emitting that much CO2 with their proompting. It’s okay that you’re pissed off, but it would be more productive if you were pissed off by real world issues.
> Pandora's box is open.
And it's not Linux' task to close that, even if it could, which it cannot. How is that disingenuous?
I agree that we should keep discussing the merits and dangers of AI-aided software development. Broad, so indeed also CO2 etc. But that discussion is not something that Linux should push forward. It's a discussion that should take place outside the context of specific projects and in the context of the AI tools and providers, IMO.
It's disingenuous because it's attacking something that the article didn't say. It didn't say "kernel devs should prompt less because it uses up water."
The real argument was that AI has many externalities that should be considered (climate change, water use, energy prices, putting people out of jobs) and that Linux, as a relevant and influential political actor, should engage with these topics. It's not even giving a final determination about what Linux should do, it's just disagreeing with Torvalds' total rejection of ethical arguments about AI.
Even if you believe that the technical merits outweigh the other costs, it seems at least reasonable to engage on those grounds. But Torvalds isn't doing that.
Also, simply stating that the "cat is out of the bag" and "Pandora's box is open" isn't really an argument. These are thought-terminating cliches; not logical arguments
> These are thought-terminating cliches; not logical arguments
Consider sometimes terminating some thoughts and replacing them with actually useful ones. It’s good for you and your goals regardless of what they are.
People burning thought-hours on pretending that we can make AI go away are basically doing nothing except emitting CO2 and consuming water. If you want change, start with yourself!
Anyway, so far zero actionable suggestions for Torvalds in the article or in this thread. Great job so far.
Contributors may object to the Hitler AI on their personal ethical grounds, but it's not going to make their buggy and vulnerable code any less so. Because our societies critically depend on projects like Linux, it follows that someone willing to pick up the pieces and do the work will then take their place.
> Sucks, but you have to learn to live with the reality that most developers will use AI to write code from now on.
No I don't. I can urge them to actually do their jobs and not just turn out slop, while refusing to use the slop machine myself.
There are many arguments one could make in favor of keeping AI around and using it, but your "argument" here is just about the lousiest and most naive argument one could possibly make, and shows a complete lack of empathy and understanding from your end.
I'm not making an argument to keep AI around or use it, I'm just saying that it stays. Reading comprehension on the internet these days...
plz dont wreck a good thing, keep ai well away from linux.
Torvalds' statement was incredibly disingenuous and rather sad. Every single commit is making a legal and political claim. Each commit is saying "this code is my original copyrighted work, and I am licensing it to the world per the requirements of the GNU GPLv2" (or "this code is someone else's work which was licensed under a compatible license, and I am integrating it into this GNU GPLv2 project").
That's not social warrior, that's simple facts. We don't need to try and guess motives or reasoning for people's contributions - that doesn't matter. What we have is an explicit declaration by virtue of the license, that the code being provided can legally be included in the GNU GPLv2 licensed "Linux" kernel. But with LLM generated code, this legal claim is dubious at best. For trivial (!= simple) code, one can probably argue that the LLM code will not be violating anyone's copyright, however for anything remotely novel, unless you are 100% certain that all of the training data was GPLv2 or compatible, it is not legally safe to use - and aside from that, even if all the training data was compatible with the license, the LLM will virtually always fail to properly attribute the source of its output, which is a requirement of - as far as I can tell - every license compatible with the GNU GPLv2. I cannot share GPL'd code and misrepresent it as my original work, I still have to make clear that it was someone else's creation.
LLM generated code is a ticking time bomb, and is going to poison projects.
Yes, it is possible to do this already without LLMs, but not by accident. A developer would have to make a conscious decision to violate someone's license in the past. Now it can happen without anyone being the wiser.
And as someone who very much uses GNU GPL licensed software because of its license, this worries me for the future of software. On technical merits, it doesn't matter, but if you care at all about the artful expression of software, as well as honesty about who created what, this should concern you.
> LLM generated code is a ticking time bomb, and is going to poison projects.
Given the recent Anthropic ruling and OAI settlements, and the general stance of courts to look the other way when it comes to LLM use, I think an exemption for LLM-based software will be handwaved into existence (to my regret.)
That certainly seems to be the way the courts are going, which is also quite troubling, considering a plain understanding of software licenses and how LLMs work should clearly demonstrate the issues I outlined above.
I don't want to accuse the courts of intentionally siding with big corporations (I have no evidence to support such a claim anyway), but the only other option I see is that the courts are uninformed on either (A) the technology at play, or (B) the requirements of many software licenses.
Though, this should also apply to works licensed under variants of the Creative Commons (with Attribution), since once again, you can end up using parts of a work that requires attribution without even knowing that it's someone else's work (or at least without knowing its license requirements). There might be a better argument for generative AI images being transformative (though I think still a weak argument since you could almost certainly produce a 1:1 reproduction of a piece of training data, much like LLMs have already been pushed to doing with snippets from books).