Much like allowing a mental patient to shave your balls with a razor.

Well there is Redox which is a Rust based OS. But even in Linux there are efforts to use Rust for certain things in the kernel and also outside. I doubt anyone wants to rewrite for the sake of rewriting but if there is code which is especially vulnerable or important for security then it's a candidate to consider using Rust instead of C.

Is it Lenovo's crappy drivers? Or all the bloatware they shove on their devices? Or the default power saving settings? Because there are a lot of things which might have nothing to do with the operating system, and a lot to other factors that happen to favour SteamOS in this case.

Physical controls don't move around, can be felt and operated by someone with muscle memory and they give positive feedback when they are operated. Haptic or touchscreen controls have to be looked at. If someone is looking at the screen or a flat surface then they're not looking at the road. While it is reasonable for automakers to remove some controls, when they start interfering with core features - indicators, headlights, volume, wipers, wing mirrors, horn, brakes, drive selector, aircon/heater, demister or anything like that they're putting lives at risk.

Tesla is the worst for it and it seems only the threat of bad NCAP ratings is making them having second thoughts. They recently reinstated the indicator stalk for example. But other automakers do not have to follow and I don't understand why they are. If the cheapest, shittiest car on the road can have physical controls, there is no excuse for anything more costly to omit them.

And all the Google stuff on Android. Unsurprising I guess that Iran has basically turned off internet to the outside world because every single phone in the country is potentially leaking information on the user.

Snowden's "revelations" were nothing of the sort. The US, China, Russia, Iran, France, UK and everyone else has and always have developed exploits and tools to compromise and spy on allies and adversaries because its in their interests to do it. This was known before Snowden betrayed his own country by revealing American means and capabilities.

But conversely in a war, it behooves countries to exercise common sense and lock down the potential for compromise and information leakage. Probably Iran should have done it sooner IMO, since Israel probably has probably compromised a lot of phones and devices used by government and military services.

Actually completely different. Since in KDE it would be KCalc, Dolphin and other nonsense names

Iran is at war with Israel and the US. It would be very stupid to be using tools provided by its adversaries no matter what public reassurances were given. Even if Whatsapp was encrypting messages end to end, it's still potentially capable of providing IP address, location and other useful meta data to an attacker.

There used to be. Now they have sensible names in the UI. So you're launching Files, Settings, Calculator etc. They might have their own names on the command line, or within their own project / community but as far as the user is concerned they have unsurprising and obvious names. Not just in English either, but the user's chosen language. This is a really low bar for any desktop.

The irony is that KDE actually has human interface guidelines and they start off reasonably - emphasizing simple by default, discoverability, least surprise etc. And then the next page actively shits on that concept by essentially says "don't separate settings / features into explicit basic / advanced". Rather it prefers you slam it all in there. So it bakes "kitchen sink" into its usability and its no wonder that users (and admins of users) prefer GNOME instead.

I've wrote comments about KDE I wrote 20 years ago and they hold true now. It's still a kitchen sink mentality and for somebody who wants to use a desktop to "do stuff" it's just needlessly complicated experience. That doesn't mean dumbing it down to the level of GNOME, but it does mean looking at the UX and asking hard questions such as - "do we really need this mess of preferences, buttons and menus on this tool when no other desktop has remotely as many". But it never happens and KDE remains mostly an irrelevance.

KDE is and always has been about throwing the kitchen sink into a Windows like desktop but never refining it or making it actually nice to use. It's just a mess of settings and menus and buttons and apps starting with 'K' or entirely meaningless names. You'd think calling the file explorer "File Explorer", or the calculator "Calculator" would be low hanging fruit in a broad, concerted effort to polish the experience but apparently not. Of course it needs much more than that - shedding a lot of the more esoteric options and general UX cleanup.

It's no wonder KDE has never supplanted GNOME as the most popular desktop for Linux and probably never will because the penny never drops.

We also need a SCO Unix version. To go with the statue of Natalie Portman, naked and covered in hot grits

So the Meta AI app is bad, but this shit will appear everywhere and in places where it serves no purpose. I have an uncle-in-law living in South Africa who came over to Ireland and had a problem with Whatsapp on his phone. The problem was it had a frigging AI chatbot stuck in the middle of the UI. He's hard of seeing so his phone had font zoom on everything and this thing was eating 20% of the screen. As far as I could tell there was no way to turn it off or hide it from the UI either and I tried.

Why did he have it but my Whatsapp didn't? No idea, but presumably they forced it on some regions with lesser privacy regs with the intent of rolling it out everywhere in due course. So enjoy AI dogshit everywhere, even in apps where it makes no sense.

Erm no, because humans reason, i.e. feed scenarios into their thought processes and evaluate outcomes. And they are affected by a greater manner of inputs and a wide scope of context than just some sentence. An LLM is basically a crank handle - same input token == exact same output token. LLMs attempt to mitigate with randomization of output (e.g. picking a token randomly based on statistical likelihood) but it's a simulacrum, nothing more.

They really aren't doing more than I said. LLMs are trained on data in a way that given any set of input tokens, deterministically it will produce the exact same set of outputs. To mix things up, models will use a "temperature" parameter that will randomly select the next token from the list of most likely outputs so it appears more random than it is otherwise. If the temperature is too high, or the model is insufficiently trained, the response is garbage. If the temperature is too low the response is boring and the same each time.

More modern LLMs might also have callbacks to allow the implementer to inject context into the response, but I'm talking about the general mechanics of what is going on.