I hate to admit it, but he’s right. I’m no fan of using AI either, but there are times and places where it can actually be a useful tool when used correctly. What’s really important is the quality of the contributions made to the project, and the usefulness should be judged on its merits alone of those individual contributions, not the “icky“ feeling people get from the idea of using AI at all.
Torvalds is clearly not here defending “vibe coding”, and he’s perfectly happy for people to not use AI at all. But when AI, as a tool, is used to responsibly, it can clearly be quite useful. And for the coders who can use it in a responsible and useful way to make meaningful contributions, I also think they should be permitted to.
He mentions the economic and environmental aspects of AI but I think dismisses them a little too quickly. While AI as a tool can clearly provide many benefits, the external costs (i.e. costs not directly born by the project but incurred by others/society as a result of use of the tool) should not be ignored.
And I don’t see mention of the seeming expropriation of works without compensation by the AI developers and the as yet unresolved IP aspects of using AI. These are part of the big picture economic impacts but also have moral and possibly legal implications.
In short, I think there are reasons not to treat AI as merely a tool.
I think its really hard for something like the Linux kernel project to look at AI from an economic and environmental aspect that doesn’t start asking hard questions about Linux itself.
AI for instance runs very well on Linux and almost certainly all of these hyperscalers and compute centers are running Linux in some form.
If Linus is going to take a hard stance on economic and environmental reasons, is he also going to ban these orgs from using Linux?
I don’t really want to do too much “what about” isms because I think it would be hella based to see the defense industry, oil industry, and AI industries all get a “fuck you figure it out on your own”. I’m just saying I can completely understand why Linus might not want to play that game since his greatest concern is to simply make sure Linux is getting better.
I agree with everything you say. All the big picture issues are difficult to impossible to understand and predict the outcomes of.
Coincidentally I saw Your AI Is Not a Tool today. I don’t agree with all it says, but I do agree with the basic premise that AI is more than just a tool because it is almost certainly going to profoundly impact society and the environment, with much of that impact as yet unpredictable.
Of course, AI will change the world whether it is used in Linux kernel development or not. And some of the changes will be and already are for the good. But I think it is more than ‘just a tool’ in the sense that a compiler is just a tool. It will be so impactful, with such significant externalities, that it is reasonable to discuss what the limits of acceptability might be, even if the stance remains laissez-faire for the moment.
Yeah he’s smart and he has the boring take which is probably the wisest take.
The wisest takes often are. He also has the benefit of decades of experience, dealing with trends and fads and upstart programmers with strong opinions.
Agreed. Every major AI model is trained off of data gathered without consent. AI even violates the licenses of the vast majority of open source projects due to the requirement that the original authors be listed in the copyright section when you pull code from a project. AI generated code does not do that.
I hate to admit it, but he’s right. I’m no fan of using AI either, but there are times and places where it can actually be a useful tool when used correctly. What’s really important is the quality of the contributions made to the project, and the usefulness should be judged on its merits alone of those individual contributions, not the “icky“ feeling people get from the idea of using AI at all.
Torvalds is clearly not here defending “vibe coding”, and he’s perfectly happy for people to not use AI at all. But when AI, as a tool, is used to responsibly, it can clearly be quite useful. And for the coders who can use it in a responsible and useful way to make meaningful contributions, I also think they should be permitted to.
He mentions the economic and environmental aspects of AI but I think dismisses them a little too quickly. While AI as a tool can clearly provide many benefits, the external costs (i.e. costs not directly born by the project but incurred by others/society as a result of use of the tool) should not be ignored.
And I don’t see mention of the seeming expropriation of works without compensation by the AI developers and the as yet unresolved IP aspects of using AI. These are part of the big picture economic impacts but also have moral and possibly legal implications.
In short, I think there are reasons not to treat AI as merely a tool.
I think its really hard for something like the Linux kernel project to look at AI from an economic and environmental aspect that doesn’t start asking hard questions about Linux itself.
AI for instance runs very well on Linux and almost certainly all of these hyperscalers and compute centers are running Linux in some form.
If Linus is going to take a hard stance on economic and environmental reasons, is he also going to ban these orgs from using Linux?
I don’t really want to do too much “what about” isms because I think it would be hella based to see the defense industry, oil industry, and AI industries all get a “fuck you figure it out on your own”. I’m just saying I can completely understand why Linus might not want to play that game since his greatest concern is to simply make sure Linux is getting better.
I agree with everything you say. All the big picture issues are difficult to impossible to understand and predict the outcomes of.
Coincidentally I saw Your AI Is Not a Tool today. I don’t agree with all it says, but I do agree with the basic premise that AI is more than just a tool because it is almost certainly going to profoundly impact society and the environment, with much of that impact as yet unpredictable.
Of course, AI will change the world whether it is used in Linux kernel development or not. And some of the changes will be and already are for the good. But I think it is more than ‘just a tool’ in the sense that a compiler is just a tool. It will be so impactful, with such significant externalities, that it is reasonable to discuss what the limits of acceptability might be, even if the stance remains laissez-faire for the moment.
These are completely valid points absolutely worthy of discussion, and I fully agree with both of them, FWIW.
I was just responding to the point that Torvalds was making here.
Yeah he’s smart and he has the boring take which is probably the wisest take. It’s just a tool and the people who vibe code just use it negligently.
The wisest takes often are. He also has the benefit of decades of experience, dealing with trends and fads and upstart programmers with strong opinions.
Also an understanding that we live in a boring universe.
The problem is that it’s overwhelmingly not used “correctly”.
Yeah, exactly. Basically the whole thing incentivizes you to NOT use it correctly. Using it incorrectly is the default.
I would say it’s more than just an icky feeling. It’s a justified moral opposition. But otherwise I agree. Just a tool.
Agreed. Every major AI model is trained off of data gathered without consent. AI even violates the licenses of the vast majority of open source projects due to the requirement that the original authors be listed in the copyright section when you pull code from a project. AI generated code does not do that.