178
submitted 19 hours ago* (last edited 19 hours ago) by usdeurpes@lemmy.zip to c/linux@lemmy.ml

It seems Linus's position is very clear now

top 50 comments
sorted by: hot top new old
[-] TeaWithDani@lemmy.world 27 points 9 hours ago* (last edited 9 hours ago)

Linux Foundation has financial backers. He's gotta give them lip service to keep them excited and happy. Anthropic is now a Gold Member. Oracle, Microsoft, Google, Meta, Amazon are all Platinum Members.

FOSS has always been subject to corporate meddling in their governance. There's a reason we call it Linux and not GNU-Linux. GNU gets no money so GNU goes nowhere.

Linus gives a headline using agents, Linux gets millions to continue doing real work. That's the game they are playing, for better or worse.

[-] wuphysics87@lemmy.ml 61 points 15 hours ago

My issue isn't with using AI to work on the Linux Kernel. My issue is working on the kernel with proprietary AI which enables Sam Altman to claim Millennium Prizes.

[-] yogthos@lemmy.ml 15 points 9 hours ago

I completely agree, using an open model would be far more inline with the spirit of open source.

[-] atomicbocks@sh.itjust.works 20 points 15 hours ago

Yeah, I’m quickly seeing a future where stuff like this AI rewrite of the Adobe suite may need to be distributed as the model that makes the software and not the software itself. If that becomes the case then we really will have entered into a world where tech bros control everything as you will quickly find yourself in a world where you can’t use the software if you don’t have the compute to make it.

[-] yogthos@lemmy.ml 9 points 9 hours ago

I can see a completely different kind of future where local models that run on commodity software get good enough for what most people need to do. At that point we'd be in a world that Stallman envisioned with every user having full control over what's running on their machine, and being able to build whatever features they want. We already see how it's becoming trivial to reverse engineer complex apps like Photoshop, or to crack PS5 games using LLMs. We are likely end up in a world where closed source is dead because code itself no longer has any value because anybody can copy and reproduce any application trivially.

And we are already starting to get there. Qwen 3.8 27b is comparable in terms of capability to Opus 4.6 which was a frontier model just half a year ago. If Qwen 4 gets to about Opus 5.5 level sometime next year, that's going to be good enough for making vast majority of software that people need. After that it's just a matter of further optimization or hardware getting better so it can run on most computers. But just to reiterate, it's already running on consumer grade hardware, and it can do real work today.

[-] atomicbocks@sh.itjust.works 4 points 8 hours ago

I’m not sure I agree with the idea that the necessary level of compute will be made available to the average person. We are already well into a paradigm where the average “PC” is an iPad or similar, with Google locking down Android and Microsoft… being Microsoft, I just don’t have any optimism that the future will have the same kind of access to personal computing power or even software access. The “tin-foil hat” part of my brain says that the purchasing years worth of chips output is part of a conspiracy to keep consumers and potential competitors at bay more than it’s a requirement for new models.

I hope I am wrong, but I am not holding my breath.

[-] yogthos@lemmy.ml 2 points 6 hours ago

There are already chips being developed specifically for running LLMs, so it's almost certain this will become akin to having a GPU in the future.

And this tech is now being developed in China as well, so you might not have to rely on western hardware either going forward. Chinese companies are building stuff on top of RSIC-V like the Alibaba chip I linked above, and I fully expect there will be a lot of focus on Linux as China continues to decouple from the US tech stack.

[-] kerryh8er04@sh.itjust.works 2 points 8 hours ago

It looks like many AI skeptics have been having a change of heart and maybe some growing dread. You spend the time and it streamlines through the tough problem and it cracks firmware, builds valid specs, reverse engineer enterprise software, it’s been an interesting week.

In some ways I think it might be good, perhaps a significant decrease in proprietary prison cage ecosystems. But at the same time, I have an ever growing dread that we are jumping off the knowledge cliff. The most experienced people are seeing revolutionary productivity, while the least experienced are gaining more capability without the foundational knowledge.

The future may have students who learn nothing, know nothing, and worse they will be happy with nothing.

I’m ready to unplug and live a near Amish lifestyle. Fuck modernity.

[-] yogthos@lemmy.ml 1 points 6 hours ago

How the education system adjusts to this tech will be interesting to watch. I expect that humans who understand how stuff works is going to be a requirement for the foreseeable future, and societies will inevitably adjust. There's always been panic and chaos every time new a disruptive technology has been invented, but humanity inevitably adjusts to it. I don't see any reason to think this will be any different.

[-] kerryh8er04@sh.itjust.works 1 points 4 minutes ago

I’d say there is something to panic about. Look at the hand made lace economy vs the Industrial Revolution, the outcome is nearly the complete loss of a skill or art form. Now take what the trend is with these LLMs, we are outsourcing the hard thinking, the struggle, the thing that forces us to expand our knowledge to a device that can do all of that 90% faster; what will the outcome be? Sure there are some people making lace by hand, but as a hobby, will we still have people in the future who will read text and write letters on paper as a hobby in the future?

load more comments (1 replies)
load more comments (2 replies)
[-] stsquad@lemmy.ml 122 points 17 hours ago

The reactions to this have been pretty hysterical considering this is an experimental branch on his own repository. I'm not surprised he might experiment with using AI agents to solve a particular pain point to see how it does. This is using agents to experiment with things that you might not have time to address normally.

Another recent example is the build system where the author said:

An LLM was used to first determine where the bottlenecks were then to figure out how to improve them. It generated a lot of code, much of it hideous.

I extensively audited and rewrote a lot of it, and heavily edited commit messages, the cover letter and comments.

off-loading the boring grunt work of identifying bottle necks and trying new approaches seems like an ideal task for a junior agent to deal with as long is you don't YOLO the result to the lists for other humans to de-slopify for you.

[-] Serinus@lemmy.world 55 points 16 hours ago

as long is you don't YOLO the result to the lists for other humans to de-slopify for you.

(Or not de-slopify at all.)

There's literally no one I trust more to handle llm code more than Linus.

He's not going to pass off shit in his name. He'll use it in the best possible way.

Lots of reasons to be anti-AI, but it's not all black and white. It's here to stay and it's useful. (But we ain't gonna spend 400 billion a year on it.)

[-] TrickDacy@lemmy.world 33 points 16 hours ago

People cannot seem to grasp that there is a difference in using LLMs to experiment and just blindly using code written by an LLM. You'd think they would realize Linus is never going to do the latter. Binary thinking leads to this...

[-] bjc@scribe.disroot.org 11 points 14 hours ago

is there a difference in the environmental destruction when used this way? no? these things have so many negative effects on the world in so many areas that making life more convenient for coddled computer touches doesn’t excuse literally any of it

load more comments (12 replies)
[-] Croquette@sh.itjust.works 6 points 13 hours ago

LLM companies will legitimize their shit by saying "look, Linus is using our model to make Linux better".

load more comments (3 replies)
[-] victorz@lemmy.world 15 points 15 hours ago

For me, there's more to it than just the output. But I'm sure Linus won't let his mind rot in LLM. Yet. Still... Lots of reasons to object to using AI at all. Use a computer responsibly, folks.

load more comments (2 replies)
load more comments (3 replies)
[-] atomicbocks@sh.itjust.works 41 points 15 hours ago

AI like this is very quickly becoming the One Ring in my mind. Like there’s a bunch of people who have convinced themselves that they can use this plagiarism box in a good way, that the outcome will somehow justify the means. But the reality is that this machine was created by evil for evil reasons and as it stands right now you can’t use Claude or ChatGPT or any of that shit and not participate in the plagiarism, in the outright theft of all of the art and creativity of humanity, and the destruction of our environment.

I wish that more people would understand that the jump from the novelty of Watson beating Ken Jennings on Jeopardy to here wasn’t technological innovation. It was purely capitalism.

[-] GamingChairModel@lemmy.world 39 points 13 hours ago

plagiarism

outright theft

It frustrates me to see that people are suddenly protective of some kind of intellectual property right of people to control creations after they've been created.

I'm pretty skeptical of AI and have my doubts that it will be a net good, but I'm definitely not gonna suddenly be pro-copyright as part of that fight. Copyright has traditionally been used to entrench corporate interests, and strengthening those protections isn't going to suddenly win a fight against some kind of AI dystopia.

[-] AAA@feddit.org 19 points 11 hours ago

People are not "suddenly protective of some kind of intellectual property". Instead they wonder why some companies can violate those rules on an unprecedented scale whiteout consequences.

[-] jumping_redditor@sh.itjust.works -2 points 6 hours ago

the reason they can "violate" the rules is that violation is defined loosely enough that feeding millions of legally aquired contents to the AI is probably as legal as making a collage.

[-] TheAgeOfSuperboredom@lemmy.ca 29 points 12 hours ago

I think it's more about the unequal application if those laws. People have had their lives ruined for sharing an mp3. These companies infringe on literally everything and they get rich.

I bet the next step is that the folks vibe coding emulators and replacements for major software suites are going to get serious fines, while artists are going to continue to get ripped off in the name of "innovation".

[-] atomicbocks@sh.itjust.works 7 points 10 hours ago

This is very much the problem I have. We have officially reached the point where the average artist on Deviant Art has less intellectual property rights than the average even LLC let alone megacorp like Disney.

load more comments (6 replies)
[-] juipeltje@lemmy.world 7 points 12 hours ago

Well i think the issue is if licenses become worthless then that's also bad for the GPL

[-] communism@lemmy.ml 5 points 12 hours ago

Exactly. What happened to "restriction of what a user can do with software or its source code is an infringement upon the users' rights"? I'm not saying that AI companies are acting in the spirit of FOSS, they're not, their models aren't even open-weight (for the most part), but the issue is not the training on materials that ought to be public domain anyway. There's plenty of ethical issues with modern technology. It's not sustainable at all, the supply chain is full of slavery, it generates huge amounts of emissions, etc. People's primary concern being pearl clutching about "plagiarism" shows where their priorities lie. If AI companies have plagiarised then I couldn't care less. It's everything else they're doing that's the problem.

[-] floquant@lemmy.dbzer0.com 9 points 12 hours ago* (last edited 12 hours ago)

I am not sure it was initially created for evil, I do believe that its discovery was mostly academic and in that it was, and still is, a pretty significant/interesting one. It was very quickly and very strongly made into the perfect weapon for the information age for sure, but i doubt that the original researchers wanted that, or just really thought of the implications outside of information theory. But it was indeed Google, surveillance capitalism incarnate, that initially gifted transformers to the world and kickstarted the whole thing...

Edit: I realised that I inadvertently made this comment about "AI" by way of pronouns when I'm really talking about LLMs. Google of course did not invent AI, and I more strongly disagree that AI in its actual sense was invented for or is inherently evil.

[-] SCmSTR@lemmy.blahaj.zone 3 points 9 hours ago

It was already clear. Anybody speaking otherwise has already been tagged and flagged as such.

[-] uszo165@futurology.today 12 points 14 hours ago

How much has he dumbed down his capabilities already? When will he replace himself by an LLM?

load more comments (1 replies)
load more comments
view more: next ›
this post was submitted on 12 Oct 2026
178 points (93.6% liked)

Linux

68054 readers
731 users here now

From Wikipedia, the free encyclopedia

Linux is a family of open source Unix-like operating systems based on the Linux kernel, an operating system kernel first released on September 17, 1991 by Linus Torvalds. Linux is typically packaged in a Linux distribution (or distro for short).

Distributions include the Linux kernel and supporting system software and libraries, many of which are provided by the GNU Project. Many Linux distributions use the word "Linux" in their name, but the Free Software Foundation uses the name GNU/Linux to emphasize the importance of GNU software, causing some controversy.

Rules

Related Communities

Community icon by Alpár-Etele Méder, licensed under CC BY 3.0

founded 7 years ago
MODERATORS