Shitty corporations are a nigh universal constant, not something exclusive to America. That’s American exceptionalism talking.

Unfortunately “AI,” and LLMs in general are, at present, wholly inseparable from the people who use them, and especially, (and more pressingly), the people corporations that make them. This is the reason I maintain my opposition to “AI” (and LLMs in general) despite machine learning as a field and LLMs as technology being truly fascinating, and having immense potential.

In my ideal world the corporations pushing “AI” wouldn’t exist, the datasets and recipes would be open, and model training would be a community led effort with Folding@Home style distributed compute used for training, pretty much removing the viability of “AI” companies. Without “AI” companies pushing for adoption, strong social pushback can be used to reign in the slopification of the commons. With those two together (and likely many other smaller changes elsewhere) LLMs can be seen for what they really are and their real pros and cons evaluated.

But that world isn’t currently feasible, and probably won’t ever be. Neither is toppling even a single corrupt corporation on the level of Alphabet, or Microsoft. So all we can do is evaluate whether we consider certain software’s use of “AI”/LLMs is acceptable, and advocate for not using them. Which is why I think efforts like openslopware, and pages like this are a positive thing, even if I don’t necessarily agree with some of the specifics.

Yeah, the same person who wrote the article I linked above also wrote one on zram vs zswap, which I think is also a good read: https://chrisdown.name/2026/03/24/zswap-vs-zram-when-to-use-what.html

Please, for the love of who ever you hold in high esteem, use swap. See: https://chrisdown.name/2018/01/02/in-defence-of-swap.html Now I don’t really like the whole “Heres an article from some rando to prove my point” thing, but anecdotally, the article does line up with my experience (with 32 gigs of system memory). Namely, when I went without swap I experienced frequent system lock-ups, when I added swap on zram, lock ups still happened but less frequently, and when I eventually changed swap on zram to disk swap on the recommendation of this article lock ups never happened again.

I don’t believe your analogy holds water. I believe computers are a lot closer to steel than “AI” is, in that they’re both foundational technologies with a very very wide array of proven uses such that the benefits of the development and use of steel/computers far outweigh the associated costs. Weapons/“AI” on the otherhand are both specific, narrower applications of steel/computer technology with a few particular beneficial uses, but those uses pale in comparison to costs of the existence and development these technologies, so much so that both weapons and “AI” should be put on a metaphorical shelf out of reach until we as a global society can grow up to develop and use these technologies in a more responsible and conscientious manner.

As much as I hate “AI” myself, rhetoric like this does more harm to movements resisting “AI” than to help them.

I used KDE for my forays into Wayland.

Bear in mind that each desktop environment has its own implementation of Wayland, meaning each can offer vastly different user experiences

Funny you mention that, as that’s (at least closely related to) one of the things I was alluding to when I mentioned design deficiencies in Wayland. The way I understand it is each individual Wayland compositor is free to create their own protocol for stuff like screenshots, and then the Wayland team will take the parts of those protocols and then combine them into the Wayland protocol for e.g. screenshots, which compositors then go back and replace their protocol with that one. Which sounds great on paper, but it leads to a problem I’ll call Ecosystem Lock-in, I don’t quite like that term, I don’t think its fully accurate, but it’s all I can think of for now. What happens is in order for a screenshot program to work on a given compositor they have to interface with that compositor’s screenshot protocol, and it’s a lot of work to support every compositor under the sun, so most devs just won’t. (Which is fine, great even, I think devs these days try to support too much stuff). But say you don’t like Spectacle (Spektacle? I don’t remember), your options are limited to whatever supports KDE’s screenshot protocol.

Or say you’re just moving from an X11 based DE and giving Wayland a try. In the process of moving over you discover your nifty little scrot based screenshot script no longer works. After some research you discover that scrot is X11 only, and the Wayland equivalent is grim, so you get your scripts converted over but they still don’t work. More research later you learn that grim doesn’t support KDE’s screenshot protocol, and Wayland compositors basically get to do whatever they want in terms of protocol design/implementation. This is what happened to me, and caused me to give up on Wayland.

I’m still using xfce; after running into problems with Wayland the three times I tried it, I decided Wayland isn’t really for me. (In fairness one of those was nvidia being nvidia, but the other two are what I consider design deficiencies in Wayland itself)

Well that explains the strange bug I started encountering in xfce recently. I guess I’ll be spending tomorrow researching DEs, though I’ll probably end up just using TDE or maybe CDE.

EDIT: clarified that the bug is in xfce, not xfwl.

Your use of Luddite as a pejorative outs you as willfully ignorant of both the matter at hand and the history of the Luddite movement. Your alignment with the corporate narrative here makes you sound like far more of a republican than any “AI” critic I’ve ever seen. Especially that “foreign operative” call out, that’s peak republican right there.

Honestly, same. I initially dismissed Zig due to their insistence on making unused variables an error even in debug builds but if they keep this up I may have to just Get Over It.

It is, just not in Arch Linux itself, which is what your comment is saying.

Minor clarification: your Arch Linux link is for archinstall, the easier install script, not Arch Linux itself. IIRC it’s not even the officially recommended way to install Arch.

Not necessarily, this only implies that computers are a part of a set of things that can be neither spiteful nor horny, and things that can be neither spiteful nor horny must never make art, not that all things that can be neither spiteful nor horny are computers.

Unfortunately most people don’t really have a choice in the matter. It’s sites like twitter that crunch images to hell and back on upload that choose for us.

There are valid criticism to be made about cosmic desktop, like

*Lists two positives*

If that’s the best example of criticism to be had about COSMIC, then it’s practically flawless. Except of course, though, it isn’t. Aside from the general lack of polish, I’d argue that there’s not enough customization. E.g. things that should really be a slider or spin control (or better yet multiple sliders/spin controls) like corner radii are multiple choice button things (not to mention that the ‘square’ style is anything but.) Some of the sliders that do exist (particularly the size ones) have very few set points that make them essentially disguised drop downs. I could probably find dozens more things to criticize if I cared to sit down and nitpick everything.