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.
back
3 comments
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!
Well thanks, but I prefer to have airplanes, and maybe occasionally even golf courses. As for climate change - you'd need to protest against Chinese (and Western, but mostly Chinese) coal and gas power plants way way before than you protest against airplanes and golf courses. I don't see too many protesters holding signs like "stop Chinese coal plants". For some reason.
Maybe because their ability to influence coal power generation policy in China is much less than their ability to influence it in the US?
Or because Chinese AI has enabled a complete crackdown on dissent?
Hmm, imagine that!
Well I see much less protests against gas and coal powered stations in the US than against, say, nuclear.
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.
Only insofar as you were able to keep a job. If you lose your job to automation, and there are no more jobs left, what do you do? Where do you get the money to keep your standard of living?
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.
> AI is not the same as aluminum cans.
True. AI is actually useful.
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?
> It argues that maintainers should consider political/moral/ethical arguments around climate change, effect on water supplies, energy prices, etc
Is the assumption that maintainers did not consider those things?
Imho, this is the crux of the debate.
Of course Linus realizes linux development is political, because he's lived that, however he might phrase it.
What I think both are actually saying:
Drew Devault (blogger): Linux development should place political/moral/ethical arguments in primacy (or at least more important?)
Linus: Technical reality should always come first, with other considerations secondary
If the above read is correct, I'm sympathetic to Linus' position. If the linux kernel doesn't work and improve, technically, then nothing else about the project matters.
Bikeshedding about political/moral/ethical qualms impairs the technical work.
Ergo, a strong lighthouse statement from Linus that the technical work always comes first, even if other considerations are made when appropriate.
The AI genie (referring broadly to 'any LLM-based tool that performs a useful development function') isn't going back in the bottle, and linux sticking its head in the sand for ethical reasons will harm linux.
One could place political/moral/ethical arguments in primacy and still conclude that the greater good is best served by continuing on.
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?
Your comment strongly implies AI is responsible for 50C temps.
AI is at the top of an industrialization iceberg going back a century, with uneven countries as beneficiaries and externalities.
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?