Thankyou, you should be an editor on Slashdot. It would be better now than being an unpaid slave doing their job for them.

Not quite. The idea is still something you have vs something you know. Passwords even if they are stored on your authenticator device as secured and not automatically transmitted over the network, unlike passkeys / MFA.

Even if both are stored on your device it still presents two very different attack surfaces that must be compromised.

Except that the passwords aren't "online". Authenticator runs locally on the device and is secured by whatever hardware security system is in place. There may be a sync function between devices, and an encrypted backup stored online, but what is being discussed is actively not "online only".

In addition to that Microsoft Authenticator supports multiple fallback systems, including typing in an offline stored single use code.

Dude this has happened multiple times, and it is super trivial. This isn't some theoretical attack. This is actively exploited by common criminals. Shit man Veritas on Youtube did it to Linus Tech Tips only a few weeks ago to demonstrate how absolutely trivial this is. It's a very simple social engineering hack.

No business has 100% uptime. Today it's cashless systems, tomorrow it's a power outage preventing you opening your register. Close up shop and open again tomorrow. If you can't survive the day without business then you were going to go bankrupt shortly anyway, and your customers will live without a sandwich.

You ignore your own generation. Those who didn't grow up with trackers in their pockets none the less grew up with email inboxes full of spam and advertisements and guess what happened... nothing. The world has gone to absolute shit, there's few things I could care less about right now than some advertising fuck send me a promotional email because they know I bought milk and a dildo last week.

Presumably do what desperate people do, and go on a murderous rampage against the government with the problem solving itself. Seriously the government can fuck you up in countless ways, pretending that cash vs card is some issue is hilarious. If you piss of someone in the government to that extent, your line of credit is the least of your worries.

Nothing, because I live in a place with data protection laws, and I am a man of the 90s and have seen my inbox flooded with spam of all sorts, and it has had no effect on me so far. I frankly don't give a fuck if someone tries to sell me something, hasn't worked in the past 30 years.

So like losing your wallet only not as bad since you still have your ID card and public transport card?

Same reason SteamOS is a distribution. Linux gaming even now in 2025 requires a lot of finely tuned parts working in unison. If you think it's like Windows where you install Steam and call it a day, you're going to have a very bad time.

Managing compatibilities between various compatibility layers like Proton, drivers (still a shitshow to install on a normal distro thanks in part to the anti-binary anti-open source limitations), and then tuning it all so it works somewhat cohesively (no end user likes spending hours in a text editor messing with config files), is not an end user friendly option.

The path of least resistance here was taking a distro, tuning it up once, installing your package on it, and then pushing it out the door, self contained, as a distro in its own right. That is the unfortunate state of Linux Gaming.

Actually in part it's also the state of Linux desktop apps, but when they tried addressing that with Snaps and Flatpacks people lost their shit because it simply is not a nice solution.

You remembered your history wrong. The devil is in the details. The question wasn't one of remote support it was one of network transparency. And the answer from the Wayland team is that no modern desktop provides network transparency over X as that fails to support even basic hardware acceleration. They said (and they are right) that when you attempt to do remote X session it does little more than render on the server and send bitmaps to the client.

So yes they were 100% right.

False. They never said they will implement it. They said they flat out wont. They said "It's a client problem". They told people if you want it, go implement it in the client. Then to shutup critics they demonstrated a basic Weston client with RDP support. But network rendering was never a feature they were going to implement in Wayland and has never showed up on any timeline and was never even remotely promised. It was and is explicitly out of scope for Wayland.

STOP GASLIGHTING. Or at least do yourself a favour and remember history. Otherwise the only lie after lie being spread is from you.

AI is an algorithm. There's nothing poisoned about it unless someone specific poisons a specific model. It doesn't profile anyone unless someone profiles you for a specific model.

You're confusing a technology with a few shitstains who are abusing people making money of it. There's no reason you can't run a completely private in house AI model. There's no reason you can't make your own AI model from scratch either (though training takes some time).

Please focus your anger where it matters, right now it's like you're getting angry at a calculator because someone published a fraudulent paper on maths.

It's not completely nonsense. It doesn't help that multiple cases are running in parallel and that there's some truly stupid lawyering involved. Again the numbnuts who sued midjourney are a classic. They argued direct infringement (which it's not, AI can't make something by itself, and doesn't store works in their entirety), and even when pointing out that their books got pirated they never actually made it a point of fact in the court. It's like they purposefully didn't argue the one case they could have one. And then in literally the same week a case on very similar grounds was found against Facebook where the court ruled that only one sub point (the storing of the data was not infringement) but allowed the case to proceed on grounds of piracy. Even in the stories on Slashdot when they were posted people got those confused with each other.

It's utter chaos.

But I want to address your other point:

OpenAI are but one company producing maybe 3-4 models. There's literally hundreds of LLM models out there, many open source. My point was that the technology on principle doesn't depend on piracy, just a few bad actors are using it that way.

Given how people respond about Google being horrible every time a search engine discussion comes up, it seems a lot of Slashdoters may well be.

That's your words, not theirs. They said influencing the environment. They never said scale or significance of impact. Their criteria was if the impact was measurable compared in emissions compared to the background, and even then you admit that they could have had a significant impact in your own sentence. How are they "kidding you" if you even postulate how the answer could be correct?

Bit more thinking, bit less knee jerking, and when you make a point try not to undermine it within the same sentence.

Why? Because they predicted what would happen within the error bars originally proposed? Is the joke on you? Is that the kind of meta post you're making?