Whenever AI is mentioned lots of people in the Linux space immediately react negatively. Creators like TheLinuxExperiment on YouTube always feel the need to add a disclaimer that “some people think AI is problematic” or something along those lines if an AI topic is discussed. I get that AI has many problems but at the same time the potential it has is immense, especially as an assistant on personal computers (just look at what “Apple Intelligence” seems to be capable of.) Gnome and other desktops need to start working on integrating FOSS AI models so that we don’t become obsolete. Using an AI-less desktop may be akin to hand copying books after the printing press revolution. If you think of specific problems it is better to point them out and try think of solutions, not reject the technology as a whole.
TLDR: A lot of ludite sentiments around AI in Linux community.
I dont think the community is generally against AI, there’s plenty of FOSS projects. They just don’t like cashgrabs, enshittification and sending personal data to someone else’s computer.
I don’t see anyone calling for cash grabs or privacy destroying features to be added to gnome or other projects so I don’t see why that would be an issue. 🙂
On device Foss models to help you with various tasks.
You are, if you’re calling for Apple like features.
You might argue that “private cloud” is privacy preserving, but you can only implement that with the cash of Apple. I would also argue that anything leaving my machine, to a bunch of servers I don’t control, without my knowledge is NOT preserving my privacy.
You might argue that “private cloud” is privacy preserving
I don’t know since when “on device” means send it to a server. Come up with more straw men I didn’t mention for you to defeat.
On device Foss models to help you with various tasks.
Thankfully I really really don’t need an “AI” to use my desktop. I don’t want that kind of BS bloat either. But go ahead and install whatever you want on your machine.
It is quite a bloat. Llama3 7B is 4.7GB by itself, not counting all the dependencies and drivers. This can easily take 10+ GB of the drive. My Ollama setup takes about 30GB already. Given a single application (except games like COD that takes up 300GB), this is huge, almost the size of a clean OS install.
Reminder that we don’t even have AI yet, just learning machine models, which are not the same thing despite wide misuse of the term AI.
But ml is a type of ai. Just because the word makes you think of androids and skynet doesn’t mean that’s the only thing that can be called so. Personally never understood this attempt at limiting the word to that now while ai has been used for lesser computer intelligences for a long time.
Well not at all. What a word means is not defined by what you might think. When the majority starts to use a word for something and that sticks, it can be adopted. That happens all the time and I have read articles about it many times. Even for our current predicament. Language is evolving. Meanings change. And yes ai today includes what is technically machine learning. Sorry friend, that’s how it works. Sure you can be the grumpy drunk at a bar complaining that this is not strictly ai by some definition while the rest of the world rolls their eyes and proceeds to more meaningful debates.
Its an interesting discussion. But I disagree you have a clear cut fact.
Just because it’s a computer writing things with math why do you say it is not intelligence. It would be helpful if you could be more detailed here.
That’s just nitpicking. Everyone here knows what we mean by AI. Yes it refers to LLMs.
Reminds me of Richard Stallman always interjecting to say “actually its gnu/Linux or as I like to say gnu plus Linux”…
Well no Mr Stallman its actually gnu + Linux + Wayland + systemd + chromium and whatever other software you have installed, are you happy now??
As someone who frequently interacts with the tech illiterate, no they don’t. This sudden rush to put weighed text hallucination tables into everything isn’t that helpful. The hype feels like self driving cars or 3D TVs for those of us old enough to remember that. The potential for damage is much higher than either of those two preceding fads and cars actually killed poeple. I think many of us are expressing a healthy level of skepticism toward the people who need to sell us the next big thing and it is absolutely warranted.
The potential for damage is much higher
Doubt it. Maybe Microsoft can fuck it up somehow but the tech is here to stay and will do massive good.
So when we actually do have AI, what are we supposed to call it? The current use of the term “AI” is too ambiguous to be of any use.
Honestly what we have now is AI. As in it is not intelligent just trys to mimic it.
Digital Intelegence if we ever achive it would be a more accurate name.
What AI means will change, what it refers to will change. Currently, the LLMs and other technologies are referred to as AI, like you say. In five years time we will have made huge leaps. Likely, this will result in technology also called AI.
In a similar vein, hover boards are still known as exactly that - like in films. Whereas the “real” hover board that exists has wheels. We didn’t stop calling the other ones hover boards, and if we ever get real ones they will likely also be called hoverboards.
To be 🤓 really really nitpicky, and i’m writing this because I find it interesting, not an attack or whatever. A tongue in cheek AcHtUaLlY 🤓
GNU/Linux is the “whole operating system”, and everything else is extra. The usefulness of an operating system without applications is debatable but they 🤓 technically aren’t required to complete the definition of an operating system.
But this is also basically the debate of Linux vs GNU/Linux vs also needing applications to make a useful operating system.
Quoting wiki summary,
In its original meaning, and one still common in hardware engineering, the operating system is a basic set of functions to control the hardware and manage things like task scheduling and system calls. In modern terminology used by software developers, the collection of these functions is usually referred to as a kernel, while an ‘operating system’ is expected to have a more extensive set of programmes. The GNU project maintains two kernels itself, allowing the creation of pure GNU operating systems, but the GNU toolchain is also used with non-GNU kernels. Due to the two different definitions of the term ‘operating system’, there is an ongoing debate concerning the naming of distributions of GNU packages with a non-GNU kernel.
https://en.wikipedia.org/wiki/GNU?wprov=sfti1#GNU_as_an_operating_system
Don’t tell me Linux mint would still be Linux mint without the a desktop environment like Cinnamon. An os is the collection of all the software not just the low level code.
Linux doesnt need GNU components at all to be a functional operating system. And you wouldnt see any difference if your http server works on GNU/Linux or Linux without GNU.
On the other hand there is difference between an AI and LLM. The difference is signifacant enough to distinguish. You may mean LLMs if you talk about AI, but tbh I though you didnt. Because many people dont.
Linux doesnt need GNU components at all to be a functional operating system.
Indeed: look no further than Alpine Linux.
Alpine Linux is a Linux distribution designed to be small, simple, and secure. It uses musl, BusyBox, and OpenRC instead of the more commonly used glibc, GNU Core Utilities, and systemd. This makes Alpine one of few Linux distributions not to be based on the GNU Core Utilities.
I won’t rehash the arguments around “AI” that others are best placed to make.
My main issue is AI as a term is basically a marketing one to convince people that these tools do something they don’t and its causing real harm. Its redirecting resources and attention onto a very narrow subset of tools replacing other less intensive tools. There are significant impacts to these tools (during an existential crisis around our use and consumption of energy). There are some really good targeted uses of machine learning techniques but they are being drowned out by a hype train that is determined to make the general public think that we have or are near Data from Star Trek.
Addtionally, as others have said the current state of “AI” has a very anti FOSS ethos. With big firms using and misusing their monopolies to steal, borrow and coopt data that isn’t theirs to build something that contains that’s data but is their copyright. Some of this data is intensely personal and sensitive and the original intent behind the sharing is not for training a model which may in certain circumstances spit out that data verbatim.
Lastly, since you use the term Luddite. Its worth actually engaging with what that movement was about. Whilst its pitched now as generic anti-technology backlash in fact it was a movement of people who saw what the priorities and choices in the new technology meant for them: the people that didn’t own the technology and would get worse living and work conditions as a result. As it turned out they were almost exactly correct in thier predictions. They are indeed worth thinking about as allegory for the moment we find ourselves in. How do ordinary people want this technology to change our lives? Who do we want to control it? Given its implications for our climate needs can we afford to use it now, if so for what purposes?
Personally, I can’t wait for the hype train to pop (or maybe depart?) so we can get back to rational discussions about the best uses of machine learning (and computing in general) for the betterment of all rather than the enrichment of a few.
It’s a surprisingly good comparison especially when you look at the reactions: frame breaking vs data poisoning.
The problem isn’t progress, the problem is that some of us disagree with the Idea that what’s being touted is actual progress. The things llms are actually good at they’ve being doing for years (language translations) the rest of it is so inexact it can’t be trusted.
I can’t trust any llm generated code because it lies about what it’s doing, so I need to verify everything it generates anyway in which case it’s easier to write it myself. I keep trying it and it looks impressive until it ends up at a way worse version of something I could have already written.
I assume that it’s the same way with everything I’m not an expert in. In which case it’s worse than useless to me, I can’t trust anything it says.
The only thing I can use it for is to tell me things I already know and that basically makes it a toy or a game.
That’s not even getting into the security implications of giving shitty software access to all your sensitive data etc.
If you are so keen on correctness, please don’t say “LLMs are lying”. Lying is a conscious action of deceiving. LLMs are not capable of that. That’s exactly the problem: they don’t think, they just assemble with probability. If they could lie, they could also produce real answers.
I’ve never heard anyone explicitly say this but I’m sure a lot of people (i.e. management) think that AI is a replacement for static code. If you have a component with constantly changing requirements then it can make sense, but don’t ask an llm to perform a process that’s done every single day in the exact same way. Chief among my AI concerns is the amount of energy it uses. It feels like we could mostly wean off of carbon emitting fuels in 50 years but if energy demand skyrockets will be pushing those dates back by decades.
Right, another aspect of the Luddite movement is that they lost. They failed to stop the spread of industrialization and machinery in factories.
Screaming at a train moving 200kmph hoping it will stop.
But that doesn’t mean pushback is doomed to fail this time. “It happened once, therefore it follows that it will happen again” is confirmation bias.
Also, it’s not just screaming at a train. There’s actual litigation right now (and potential litigation) from some big names to reign in the capitalists exploiting the lack of regulation in LLMs. Each is not necessarily for a “luddite” purpose, but collectively, the results may effectively achieve the same thing.
“It happened once, therefore it follows that it will happen again” is confirmation bias
You’re right but realistically it will fail. The voices speaking against it are few and largely marginalised, with no money or power. There will probably be regulations but it will not go away.
You misunderstand the Luddite movement. They weren’t anti-technology, they were anti-capitalist exploitation.
The 1810s: The Luddites act against destitution
It is fashionable to stigmatise the Luddites as mindless blockers of progress. But they were motivated by an innate sense of self-preservation, rather than a fear of change. The prospect of poverty and hunger spurred them on. Their aim was to make an employer (or set of employers) come to terms in a situation where unions were illegal.
I think the biggest problem is that ai for now is not an exact tool that gets everything right. Because that’s just not what it is built to do. Which goes against much of the philosophy of most tools you’d find on your Linux PC.
Secondly: Many people who choose Linux or other foss operating system do so, at least partially, to stay in control over their system which includes knowing why stuff happens and being able to fix stuff. Again that is just not what AI can currently deliver and it’s unlikely it will ever do that.
So I see why people just choose to ignore the whole thing all together.
This and on top of being inexact, it’s not understandable and un-transparent. These are two of the top reasons to push for free software. Even if the engine executing and teaching models are free, the model itself can’t really be considered free because of its lack of transparency.
That is a stretch. If you try to download and host a local model, which is fairly easy to do these days, the text input and output may be semi-random, but you definitely have control over how to plug it into any other software.
I, for one, think that fuzzy, imprecise outputs have lots of valid uses. I don’t use LLMs to search for factual data, but they’re great to remind you of names of things you know but have forgotten, or provide verifiable context to things you have heard but don’t fully understand. That type of stuff.
I think the AI shills have done a great disservice by presenting this stuff as a search killer or a human replacement for tasks, which it is not, but there’s a difference between not being the next Google and being useless. So no, Apple and MS, I don’t want it monitoring everything I do at all times and becoming my primary interface… but I don’t mind a little search window where I can go “hey, what was that movie from the 50s about the two old ladies that were serial killers? Was that Cary Grant or Jimmy Stewart?”.
I think the biggest problem is that ai for now is not an exact tool that gets everything right.
The biggest problem is that it isn’t an exact tool, but is being presented as if it was and implemented as a replacement for people instead of a tool they can use to make themselves more efficient.
Good point about the imprecision. On the other hand most Linux desktop users are Normie’s, think Steam deck and so on.
Some of the most popular Linux desktops are built for ordinary people with the KISS principle in mind. Not arch using tinkerers
On the other hand most Linux desktop users are Normie’s, think Steam deck and so on.
Jesus fuck what a statement. Your parents probably regret having you.
That’s not the tone I like to read even as an answer to a statement I don’t agree with. No need to get that personal.
One of the main things that turns people off when the topic of “AI” comes up is the absolutely ridiculous level of hype it gets. For instance, people claiming that current LLMs are a revolution comparable to the invention of the printing press, and that they have such immense potential that if you don’t cram them into every product you can all your software will soon be obsolete.
