Kogasa

Fly-brain-shaped neural net with some explicitly programmed behaviors. The actual biochemistry of the brain is not modeled.

Immigrants good tourists bad

Official support is a higher bar than “we did the bare minimum to make sure it’s theoretically possible for Proton to run it.” They’re probably just doing a little testing to make sure everything works decently before endorsing their product on Linux.

sparkling arch, sparkling fedora, and diet arch

That doesn’t follow. What I think you mean is that by defining the constraints so rigorously, you’ve basically solved the problem yourself. But that’s exactly the point. The LLM isn’t the problem solver, it’s the execution engine that does the wiring-up and the ticking of boxes. The easy part, arguably, sure. But still considerable effort that can be saved, and that effort may be better spent on the problem-solving + constraint-defining stage.

Mandatory disclaimer that this is not a pro-AI post

Using aur helpers is fine if they make it easy to read the pkgbuild, which paru does. It’s too annoying to check for PKGBUILD and upstream/vcs updates for each package individually.

Ideally the aur helper would point out when 1) a package changed maintainers since your last install, 2) a package’s PKGBUILD itself changed (not just the upstream/vcs source), 3) the PKGBUILD is less than 24h old or so. And for #2, it should also show you the changes similar to what you see on the AUR site’s “view changes” page. I’m not aware of any aur helper that does these things, but hopefully recent events prompt a change.

Do harder content and it suddenly becomes a real job. Keys are mostly vibing up to +21s right now and then suddenly you have to keep track of everyone’s defensives and externals and tell them how to use them while also rotating your cds perfectly

This particular seal sounds like FWHHHH GUHH MMMGG GGUHHH but that’s splitting hairs

These sea lions are well trained and familiar with the concept of “play along and acquire fish”

They are very good boys and girls

Gullikit ones generally have this functionality

Just change your name to Gimp.

People are using LLMs to diagnose disease, write prescriptions, deny health care claims, deny loans and grants, write scientific papers, review scientific papers, draft engineering and architectural documents, and talk to their loved ones

Despair

I don’t support the change. That’s not my point. My point is that if we’re going to argue the dev being threatened isn’t a victim because he’s actively harming privacy, we should be aware that the changes he proposed are not actually harming privacy at all.

Age verification laws: slippery slope. Sure. I agree.

Adding optional age field to systemd userdb: not slippery. Systemd isn’t being weaponized as an age verification suite. It’s just not happening.

You don’t, and you don’t have to fill them in with accurate information, so it isn’t.

It’s mostly not going to be used at all.

As I said, I also object, but you have to realize you’re literally just doing the slippery slope meme unironically. The part that makes it a fallacy is the unjustified assertion that more egregious changes are the inevitable result of the first one, except the first one is materially harmless and in line with existing PII fields in userdb. It’s completely reasonable to expect systemd to go no further than it already has.

No, it literally just can’t violate your privacy in any way. You have complete control over what, if anything, is placed in that field. No information about you can be gained or disclosed by virtue of the systemd change alone. You can think it’s a bad change because it signals intent to follow a trend of supporting privacy-invading age verification, but you can’t say this specific change in itself is privacy-invading.

“user is likely accessing service from a *nix device” isn’t PII. of course anything can be used for fingerprinting, but this type of “leak” is about as insignificant as it gets. It’s not what most people would consider a violation of their privacy.

I don’t think the changes in question are “upholding” any law, but rather giving system admins and software devs a convenient/predefined way to attempt to comply with the law if they choose. “Upholding” the law would be requiring the field to be filled or checked.

That said, to your point, if someone proposed a race field “so that devs can implement segregation if they choose,” I’d find that reprehensible even though it doesn’t do anything on its own. Similarly, I object to the systemd change.