GrapheneOS is currently defending its use of AI coding tools on Mastodon against complaints by various accounts claiming to be users.
We do not understand where you’re coming from or why you’re so incredibly angry with us. It’s not justified and does not make sense.
If they’re just using it to find and patch bugs/exploits? Fine by me.
Vibe coding? Get the fuck outta here.
This is the key point.
From my reading that is not the limit in how they are using AI. It seems like they are accepting LLM generated code as inputs.
This pretty much compromises the whole project as far as I am concerned.
If it’s thoroughly vetted, it might still be OK, but it’s concerning…
There is no making ai go away, it’s never going away. Now it’s a tool to divide us.
I haven’t dug into the code, but blanket hate for AI in code is extremely frustrating. There’s a difference between vibe coding and the hundred other ways to “use AI” on a project.
It’s funny because anyone seeing this as black and white clearly hasn’t ever opened an IDE.
Usually, if it’s noticeable that AI was used on code, the code is flawed or, at least, has copyright concerns.
Agreed.
I strongly remember a CompSci professor ~2005 being disgusted with auto-generated Java GUI. And yes it was clunky and more memory intensive than it could have been, but also, I really don’t care about how GridView is defined most of the time.
Bad code is bad code, touched by AI or not. For working code, a volunteer group spending resources to determine if it might have been touched by AI is at the very least an incredibly frustrating standard to try to meet. And very likely impossible to verify in any meaningful way.
Shh, you’re using logic on social media.
If the haters only realized that literally every software company has been using it since at least last fall, we’d have a helluva lot more Amish.
Look at the comments here. Do you think the haters here care?
Noooo All AI is Baaad!!1!
/s
It’s tiresome.
People still seem to think AI is a fad that will go away. I don’t see this happening (for code) at all. Not ever
I don’t like GOS devs, they are not focused on usablity, but on security (which is completely valid, but it’s a shame that GOS is the most usable mobile OS), but holy based.
If you are disputing AI utility in coding: Recent Linux kernel versions fixed up to 3 times more CVEs in comparision to releases before AI. Also, much less of those CVEs are found in the newer versions, which means new kernel releases contain better code.
I am not a coder in any fashion but this is the make or break frontier for development, particularly FOSS and small scale projects.
AI is evil.
It is an ecological nightmare, socially and economically distructive, prompted on the theft of the knowledge and work of every creative endeavour any person as ever made and pointed at rendering pointless any future aspiration of individual creation someone may have.
The association has been made and it is not going away.
AI is a tool, like a hammer, and should be used as such. And like a tool should require training. The problem is people aren’t using it as a tool and rather look at the hammer and say “go build me a house”. That’s where the issue is. Does a hammer know how to build a house on its own? no. does AI excel at end to end builds? no. but people are using it for said end to end builds.
Devs, vibe coders, god knows who else still to this day fail to understand that AI doesn’t understand how to build things correctly. All the agents care about is getting to a solution as quickly as possible to retain a userbase. that’s it. will what ever it builds scale? no. will it be full of exploits? very likely. will it be able to refactor whatever it wrote? no because it will believe YOU wrote it. so upgrades/improvements to whatever are virtually impossible.
THEN you combine all that with the fact that EVERY AI’s training data, currently, is incredibly out of date. Tech moves too fast for it. So you have people building shit with libraries and what have you that are already out of date and they don’t know that.
It’s a tool like a gun, specifically purpose built to usher in a technofeudalist surveillance state. Not useful for anything else as is.
A tool built through nefarious means and consuming resources it should never had chanelled into it.
Then, taking from your words, if the information these machines are working is by default outdated, why insist in feeding it with more and more and expect different outcomes?
Makes no sense to begin with.
Nuclear weapons are a tool, like any other
That’s a great example actually, you just have to differentiate between a nuclear weapon and a nuclear power plant. You can split atoms to build a bomb or heat water, the concept of splitting an atom itself isn’t evil.
Exactly
I think with OP’s analogy, a nuclear weapon is no longer a tool, because the actual tool (nuclear technology) was crafted into something (weapon) that does a specific thing (destroy). But I guess OP should’ve used LLM and machine learning instead of AI.
Nuclear fission has been used in nuclear reactors for good, and was exceedingly efficient at that time, and has been a significant factor in the recent human history. Just as it’s been with the “evil” use in nuclear weapons.
I’m not aware of any breakthrough that has not been used for both good and evil.
AI in concept is not evil and could be a good thing .
Unfortunately, its implementation has been at the hands of rapacious, avaricious, and narcissistic über-capitalists, which has inextricably linked AI with the evils you enumerated.
In a different timeline, AI was developed slowly, using heavily-vetted and legally obtained data, and rolled out in an intelligent and safe way.
We are better off leaving the appeal to alternate time lines for another conversation, preferably around science fiction or comics.
As it exists, as it was created, “AI” is evil. It consumes resources we can’t afford to waste and is currently supercharging our pollution indexes and creating what may be a catastrophic collapse of already very fragile and unbalanced social and economic realities.
And because we are already into the sunken cost falacy fase, it will never be rewritten or rebuilt. Hopefully, it will collapse in such horrifying way it will create a cautionary tale, but we are not so lucky, so what we are left with is picking up the bits after it.
This kind of system had to be developed inside laboratories, colleges, universities, research centers, to aid into very niche and specific fields of research, not made to be a mass consumption comodity as it is. But it was, as you very well put it, not done that way.
AI is not evil. It doesn’t have a consciousness. It doesn’t make decisions.
What is evil, are billionaire capitalists having full control over the technology.
What is evil, is the concept of closed-source.
What is evil, is stealing all creative works without attribution or compensation.
What is evil, is seizing essentially all the silicon to build these datacenters, then soak up all the electricity out of the grid and all the water out of the ground.
The problem isn’t the AI, itself. It’s the means that’s the problem…not the ends.
To summarize: something that is being created through highly destructive amd predatory means, by completely deranged people, is not evil.
I can’t disagree more. And to quote myself, again, today:
As it exists, as it was created, “AI” is evil. It consumes resources we can’t afford to waste and is currently supercharging our pollution indexes and creating what may be a catastrophic collapse of already very fragile and unbalanced social and economic realities. And because we are already into the sunken cost falacy fase, it will never be rewritten or rebuilt. Hopefully, it will collapse in such horrifying way it will create a cautionary tale, but we are not so lucky, so what we are left with is picking up the bits after it. This kind of system had to be developed inside laboratories, colleges, universities, research centers, to aid into very niche and specific fields of research, not made to be a mass consumption comodity as it is. But it was, unfortunately, not done that way.
I am not a coder in any fashion
You should use ai to learn 😁
I recommend https://chat.mistral.ai/chat as they use renewable power to train their models and are european 👍
Pass. Haven’t learn before, not interested in do it now. And if by some reason I wished to do it, I’d grab a few books on it.
“Ai” isn’t inherently evil. It’s just a tool. An incredibly interesting and marvelous tool, in theory.
What we make of it and how we use it matters. Noone needs stupid chatbots. People should only be allowed to use them after graduating an exam of what they actually are and how they work. In the current state they’re 90% bullshit and harmful.
But LLMs surely have great and reasonable applications. People often fear what they don’t really understand. And LLMs really are a topic that often even “techies” don’t get, so why should a layman.
Yes, they are environmental Desasters. But try to substract all the trillions of silly nonsensical queries millions of people hammer into chatbots, that once were simple search-engine - queries. And it’s not THAT bad anymore.
“AI is just a tool” - #1 thing said by tools.
Machine guns are tools. As are jets and bombs and missiles.
Exactly. Seems you understood the point. Also same as cars, planes, spoons and machetes.
“AI is just a tool” - #1 thing said by tools.
#1 thing said by arrogant turds with no clue.
My point proven :D
If that fills your clueless heart with joy, be my guest 🤷🏻
I’m just going to quote myself, as I’m satisfied enough with the answer I wrote to another reply.
As it exists, as it was created, “AI” is evil. It consumes resources we can’t afford to waste and is currently supercharging our pollution indexes and creating what may be a catastrophic collapse of already very fragile and unbalanced social and economic realities. And because we are already into the sunken cost falacy fase, it will never be rewritten or rebuilt. Hopefully, it will collapse in such horrifying way it will create a cautionary tale, but we are not so lucky, so what we are left with is picking up the bits after it. This kind of system had to be developed inside laboratories, colleges, universities, research centers, to aid into very niche and specific fields of research, not made to be a mass consumption comodity as it is. But it was, unfortunately, not done that way.
You are confusing capitalism and ai. AI is just a tool. Capitalist will use any tool it can to hurt labor. This is currently born out by history recent and ancient. The luddites attacked the looms not because they were evil. But because the bourgeoisie were using them to exclude and attack labor.
That is a false equivalence.
The looms, the automated treshers, the tractor, whatever machine designed to a fullfil a task that displaced human labour wasn’t designed with the objective to render humans useless. There were periods of transition when people hurt, needlessly, if I may say so, but new occupations emerged.
This is not the same.
For the first time, we are faced with a technology that feeds off human achievement to be built and next states it can dispense and completely replace what enabled it to be put together to start. While consuming resources we can not afford to waste and returning nothing of value.
No, it’s a literal analog. There will still be people operating AI. Even using your own terminology is literally the same.
And I don’t even know where you’re going with the last part. I suggest you familiarize yourself with what AI is actually capable of. The biggest danger of AI is not that it hallucinates more than someone tweaking on meth. But that people dumb and tweaking on meth with too much power will take what it says as gospel. Your liable to make the people in the marketing Departments of these AI companies cum with how badly misrepresenting what these AI are capable of. The same AI that some companies fire their workforces for and are now humiliatedly hiring them back. Because the AI wasn’t capable of what it was sold as.
It is not an analog, in any measurement.
We are talking about a technology that been prompted, since its inception, as the leap that would render human work unnecessary - which is false under many measurements.
When that failed, it was then repackaged as the technology that would enable any person to do anything: write a book, a song, create the new Mona Lisa, write any program they needed or wanted, etc. And again, it failed.
Any major development was always geared towards solving a concrete, tangible, problem. For AI, every proponent for it is trying to find problems to fix with it.
I don’t care about the promises of AI. I care about the cost of it, now and future. It’s not a hand full of niche applications that are going to impress me. Put it towards research, make it truly useful, in the hands of people capable of forcing it to produce results and audit and validate them. It is not a toy, it is not something that can replace human beings in any measurement, it is not something people should be relegating their autonomy and individuality to, which is happening as we are having this conversation.
No need to reply to this. I’m out. I don’t care for any arguments in defence of AI and it’s potential. I’ve made my choice. Keep to yours.
Have a great one. See you around.
“I’m just going to quote myself”
Yeah go on buddy, doesn’t make you right though.
AI is an item (as “tool” seems to trigger you people), saying it is “evil” is just stupid and makes you sound like an evangelist lol.
Is it bad? That is something people can actually discuss, but you’re not here for that it seems.
The AI is a consummer of resources we can not easily dispense that returns nothing of objective and equivalent value.
Fresh words, no quotes.
More palatable now?
Oh no we didn’t know that!! /s
Dude, we know the downsides of generative AI and the training of llms.
Is that so hard to understand?
It is. It really is. At any point on any other endeavour, someone said it was enough and the costs did not justify the results. Not with this. So, it really is hard to understand.
Just because it was a nicely worded statement, doesn’t make it universal.
I use a local LLM for our smart home. It wastes a Lil bit of energy, which is still faaaar below my 1500W while gaming. Is gaming evil now? It helps dramatically in ways that were never possible before like “turn on all lights at 40% in the color purple and also turn on music. My fav playlist” followed by a “no, use a warm color and close all windows”.
Not every LLM is a fucking chatbot that eats the planet to “create” a shitty cat-meme for Granny’s Facebook.
Don’t assume everyone uses silly chatbots. That is ONE application of LLMs, and it is indeed stupid and kills out environment for basically no benefit.
Congratulations. Enjoy it.
https://grapheneos.social/@GrapheneOS Here you go, argue with GrapheneOS - it’s an open platform, go ahead and make a masto account and share your thoughts with them. Tell them what you think! The rep who handles their masto account is very well spoken and presents rebuttals of value so I advise you bring your A game.
The rep who handles their masto account is very well spoken and presents rebuttals of value so I advise you bring your A game.
Yeaaaah like deleting their earlier posts to mke themselves look better and claiming the “haters” are coming from K*wi Farms.
https://mas.to/@zzt/117246306723269935
Truly good rebuttals right here!
I’m not going to approach them in any fashion. I have already made my decision about their work/product and the GrapheneOS is not of my interest.
The project must have their reasons but I still insist on what I’ve stated regarding AI in general.
You’re right, but don’t encourage people to brigade GrapheneOS. That’s a donation-funded project.
I’m not brigading anyone. They have their reasons, I have my view regarding AI, that is it. I did not attack the GrapheneOS team in any way nor do I intend to.
My bad, I didn’t think of that. My only thought was “$20 says this chucklehead won’t do squat” because we know.
I’m with Linus Torvalds and GrapheneOS on this. AI is a useful tool, for all its flaws and negative externalities.
I am sure I will be down voted on Lemmy for this but using AI for development by a team who is already experienced in development of the product they are working on is not a moral failing.
Don’t forget the negative internalities (it makes it’s users dumber and the codebase less maintainable).
But yes it is useful, and the knee jerk hate comes mostly as a reaction to the marketing lies told by the companies selling it.
Luckily, the “negative externalities” are mostly made-up bullshit.
Reality altering disinformation as fact, using a machine that is essentially plagiarism incarnate? You are pushing a false narrative OP
The ecological damage is not made up.
It’s pretty insignificant compared to the rest of the economy. Focusing on AI to the exclusion of bigger issues is actively harmful.
It also didn’t exist at all 5 years ago. Why are we as a society allowing new ecological damage? When will we stop? The billionaires want it to be never as long as it gives them money and power.
Renewable energy makes good progress, but as you can see here, many people have other priorities.
It’s pretty insignificant
Citation needed. You’re just straight-up lying or clueless as to what’s actually going on.
It’s not hard to look up. If you need help, ask for help politely.
That’s an “I have no idea what I’m talking about” if I ever heard one.
“Citation needed” IS polite. You claim it’s not polluting, so clearly you know the CO2 emissions numbers from SpaceX’s datacenters that are being run off direct-connect to natural gas generators. So, how about it? How “not polluting” are they?
Unfortunately, as you can see, you are in a crowd of liars. Now, I am certainly not accusing you personally of being dishonest, but it does tend to bias your audience against you. Sorry.
The numbers aren’t particularly bad. It’s not for nothing that the EU has classified natural gas as green energy.
Layoffs are made-up?
Layoffs aren’t the problem here…the problem is that society requires people generate their own worth through work, and suddenly a good chunk of that work is now significantly less valuable.
Requiring less work is a good thing. Humans should be striving to minimize work. That’s what humanity is good at.
The problem comes in that while the “input” of AI is, essentially, the collection of all human knowledge…while the output is money for a small handful of people…people who themselves contributed next to nothing to that pile of human knowledge (and in fact are probably running a deficit to it).
That ain’t right. We are all shareholders in the collection of all human knowledge. We should all be getting dividends from its work.
Oh look! Another “leftist” trying to shift blame away from tariff man.
I’m not American if that is what you’re implying. Have a nice day.
Maybe I am just not understanding your direction based politics, but isn’t it the right that put him in office? Isn’t your comment backwards?
To be clear, I literally don’t know, it’s hard enough following my own country’s politics lol
Yes. Trump’s deranged acts are seriously harming the US economy but also doing a lot of damage to the global economy. Anyone who blames AI for layoffs is effectively helping Trump/the global right by covering up their incompetence and failure. To be fair, Putin’s war is also causing much harm.
Then why did you say they were leftist?
As for the layoff blame, just because I say I like hamburgers doesn’t mean I hate hotdogs. Layoffs can be happening from both those things, and just because people are acknowledging one source doesn’t mean they are helping the other source…
I said “leftist” in quotes. Almost everyone here self-identifies as being on the left, at least insofar as they oppose Trump. And yet… well.
I am skeptical if AI is already having a substantial effect on employment. In any case, I don’t think these people care. You don’t see anyone arguing for a better social safety net/unemployment benefits. Here’s someone saying that people lose their jobs over AI, but you also have people saying AI is useless. They really don’t care.
@bekkaboo@girlcock.club Linux itself is heavily using AI models.
For security analysis and uncovering bugs.
Oof. Options for good software grows smaller and smaller each day.
hey feel free to fork it, call it amish os!
I’d fork a lot of software if I had the time and skills, an OS is a bit too much for me. I’ll happily donate to good projects that come around though, I’m just not feeding money to ones that are feeding that back to billionaires to build datacenters for fucking AI.
Cool well I’ll keep using Lineage until they become massive piles of steaming crap as well.
Fuck AI and all you silicon ball lickers.
@BrickEater unfortunately LineageOS does allow AI contributions: https://github.com/LineageOS/charter/blob/main/ai-coding-assistants.md
if you are using a computer or phone, you are running AI-generated code. it is rather dire. the best you can do is limit your use of this fashtech in other areas.
just curious, are you planning to stop using technology altogether, once there’s nowhere else to migrate to?
I mean, I assume you have thrown away your PC already, since all 3 major OSes use AI in their development
No they don’t.
Windows clearly does. Linux kernel does. MacOS has built-in AI features and XCode has support for AI agents, so one would assume they also use it in development. Do you have some other 3 major OSes I don’t know about? Are you on FreeBSD? Cause uhhh https://freebsdfoundation.org/blog/freebsd-ai-assisted-vulnerability-discovery-project-launch/
Linux, Windows, and MacOS devs definitely do use LLMs to help with coding.
In your defense NetBSD does not, but I would not consider NetBSD to be one of the big 3.
A large number of people are using it as an opportunity to promote using operating systems which are the direct opposite of hardened and are much more heavily using AI models including for a lot of code generation.
LineageOS is certainly using it. CalyxOS, /e/ and iodéOS are based on it. CalyxOS is developed by LineageOS core developers.
Since LLMs are essentially built from code, I have an easier time believing that coders can get a benefit, especially at the Individual level, within the frame of having a coding background and being able to audit, directly, the code that the machine spits out. But basically everything else happening with, and surrounding, these fucking things is atrocious.
good to see smart developers posting reasonable takes, though sad they have to spend time doing it
Remind me to stay away from that Nazi bar.
Defenders of genAI might as well hoist their vintage early 1940s German flags while they’re ahead! Though I suppose an Italian or Japanese flag from the same time period would also work as well.
What has AI got to do with Nazism?
For all their faults (and there are many) the GOS dev(s) are one of the good guys.
I guess many people don’t know how much programming has changed over the last year. GrapheneOS is doing a good job telling it like it is.
It hasn’t changed at all, some of us just use LLMs to help us code.
What has changed is the influx of morons thinking they are devs cos they can vibe code an app full of bugs and security risks.
… telling it like it is.
I know only one type of person that uses that phrase unironically.
Oh yeah. Op is definitely one of those types.
Your take here is so dumb. Many on Lemmy are software engineers (myself included).
We know exactly how “AI” companies have tried to change programming and know how dumb and misguided it is because we understand quality code.
Sure bro. All dumb takes by GrapheneOS, LineageOS, and even Linux.
You understand quality code, unlike Linus Torvalds.
Appeals to authority and hero worship are terrible arguments for anything
GrapheneOS and Linus have always had a god complex and took the stance their way is the only way. I have never respected them as the arbiters of the correct path others have. (I know LineageOS exists but not much else so can’t speak to them).
I respect them putting in the effort. I respect them trying. I’ve contributed where I can because the overall goal of data privacy and digital freedom is one I agree with. But to hold them on a pedestal as always right is, yes, still a dumb take. I’ve never seen anything that made them the end all be all people seem to hold them up as. Solid voices who’s opinion should hold more than average? Sure. But that’s it.
Yes, Linus is smart and has had good takes and can produce code used by most every computer in one way or another. That doesn’t make him infallible.
Its not about putting linus on a pedestal… its about these anti-AI purists declaring everyone who uses AI a nazi and all code is slop. Big tough guys saying they wont use software with even a single line of generated code.
Go ahead, fork the linux kernel then… these people will never be able to follow through with their purism, unless they go full templeOS mode
I dislike AI overall, but what i’m starting to dislike almost as much are these anti-AI purists who are holier than everyone
I mean this in the worst way possible. Your attitude mirrors the far right and is just as bad a take. That’s their approach to everything.
“Oh, you’re fighting against something you see as a danger to literally the entire world? Well if you’d just stop shoving your opinion in our face we’d be happier.”
He doesn’t care for misguided opinions not backed with anything but emotion. The people he is talking about don’t actually look into the issues and fight those, their mind is set that AI is evil and you are too for any usage of it, without actually understanding anything about the situation.
There are situations where using AI uses less power than traditional methods, there are situations where the power is clean, hell there are even models getting made that don’t use stolen data. Some people use it more responsibly (use tiny models where possible) while some toss everything at whatever new Fable or Opus is available.
There are levels of AI usage and their blanket statements with no understanding are annoying because it alienates the people that are responsible with their usage.
I am sorry you get tossed out with the bathwater.
It absolutely is something I struggle with. To me, I know LLMs are just called AI but aren’t, and machine learning is not LLM. I’ve actually developed a neural network based program for a previous company before and know it can be useful in certain cases.
But how do I know the person I’m talking to knows that? How do I know the person is one of the minority who is trying to do it ethically and responsibly? How do I express just how truly damaging the current mainstream “AI” methods are succinctly and strongly without getting bogged down in that discussion and sidetracked from the bigger picture (which is exactly what’s happening here in this comment for example).
Ok, so i’m far-right (just call me a nazi) because i dont have a 100% blind hatred of AI?
this is exactly what i’m talking about… this kind of holier than thou attitude makes enemies out of potential allies
I’m sure if we had a reasonable discussion about AI we’d find 90% of our opinions align… i’m generally very against AI as it is now…
But instead you’ve already labelled me as just another pro-AI nazi because i dont pass your purity test
I explicitly did not call you a Nazi or far right. I do see how what I wrote could be interpreted that way and isn’t my intent so my bad on that one.
I said that that’s a far right attitude to take. Shutting out the opinion of people against it because you view them as obnoxious. Nothing good will happen if we all did that in my opinion.
fighting against something you see as a danger to literally the entire world
Is that what you’re doing, then?
Unfortunately, having a god complex is not sufficient to make one a good coder.
You’re literally the one who was using him as your example…
You’re an idiot if you think popular and field leaders can’t be wrong.
Mastodon users, yuck
They’re not using it for coding, they’re using it to review code.
One look at their agents.md says otherwise
https://github.com/GrapheneOS/Camera/commit/da7e925ebecdaa2249fabba67a6a791bb8dd384b
No, it started with them saying it was for review, but they added any of their coders can use it if they wish.
There is really no point arguing about such details. AI haters don’t care how exactly you use AI. If you want the best, most secure software possible, you need to let the developers find whatever way there. Imposing arbitrary restrictions won’t work.
No, you need to make sure vulnerabilities aren’t introduced by lazy or underpaid developers.
Its so annoying that all code has been perfect and bug free until these silly LLM people came and made buggy code
“Perfect and bug-free” <---------------> “Chock full of slop and security holes”
This is a spectrum, but you already knew that and are just trolling.
I’ll grant that bosses and managers have a valid role for keeping the work on track. But someone who tries that without being qualified themself will cause major problems.
