Everything you described sounds like a dystopian nightmare straight out of Upload.

T H E O R B

In this case it’s linux. I am prepared to pay gobs for proper linux devices. However, I’d rather something like a liberux Nexx if that thing ever comes to fruition.

Are we married to this idea?

Dot com 2.0 wooooo

Not to piss in your cheerios, but I like this new design.

The logo is kind of mid but it integrates with modern themes better than before (I will miss the old logo).

It’s an absolute blessing that they picked Qt instead of GTK and I want to see more of this.

How do you feel about the color profiles on HDR, and the cursor/stylus sync, or any of the other changes they listed in this one as far as art goes?

This is a massive drop for artists isn’t it?

To be clear, I am not arguing in favor of what I suggested. I am vehemently against it. But I have worked with the technology enough in a professional setting to know what it is capable of and how it can (and I have) used it. I am afraid of what I am seeing as a possible (probable?) future.

I’ve seen some law nerds explain it recently as:

De-jure monopoly=the one we all think of (the bad one). This comes about through dirty tactics, anti-consumer practices, and backstabbing.

De-facto monopoly=the one Valve is (and not illegal). This comes about naturally as customers independently choose one company over all others for nothing other than they like them better.

Librewolf is pretty legit.

If an AI can tell what’s in a picture (which is a series of pixel data), it can certainly reconstruct gcode into an “image.” Ive used it at work for similar tasks.

Cylinders obviously are too generic, but there are parts that are a lot more specific and could be fingerprinted.

And now two! Touchpads work like a dream for me

Not sure how your clipboard is broken, it works using KDE connect. DeskFlow may need to work on implementing the Wayland bits.

You’re gonna hate this, but… AI can literally do it, and for the large models it’s terrifying how accurate they are. You will argue that your little ESP32 powered reprap or klipper or whatever printer can’t handle it, to which regulators will go ok then, either the printer has to call out to a service with an http request to upload the gcode every time it wants to print anything, or your slicer has to do it (and we dont care that it’s open source, it’s illegal to operate if it doesnt make the call and you’re getting fines or jail time if you get caught).

This is what AI was built for 😟

EDIT:

In case it isn’t abundantly clear, I am not in favor of what I just described. I know that it is possible though and know how to architect exactly these mechanisms. If I can build them, so can they. (I won’t, of course, that goes against everything I stand for).

I wish GOG supported Linux at the level Valve does. On paper it sounds like a match made in heaven. But they don’t, and the strange things they do to the executables of old games to preserve them makes them difficult to get running under wine and proton smoothly.

I may start keeping a cheap device that lives in a Faraday cage that obeys the corporate rules and only comes out when I absolutely need it, and then a graphene device of sorts as my daily driver. Ive almost completely de-googled otherwise.

Is it?

As a malicious actor or red-team player, I would want to get you on as old of an OS as I could in order to exploit a wider range of CVEs. Or in most cases, one would be hunting for a specific set of CVEs. Once I’ve got you on the version I want, I can then perform other attacks and ensure that they run.

The iPhone, many Android phones, some network equipment, and game consoles all have eFuses that burn when you perform an update, and the specific number or pattern they burn in is used to determine the lowest OS version your device is allowed to be on in order to stop this from happening.

Thats called a downgrade attack and is explicitly blocked by most modern security models that are not a PC.

Do you think it runs at 1000w continuously? On any decent GPU, the responses are nearly instantaneous to maybe a few seconds of runtime at maybe max GPU consumption.

Compare that to playing a few hours of cyberpunk 2077 with raytracing and maxed out settings at 4k.

Don’t get me wrong, there’s a lot to hate about AI/LLMs, but running one locally without data harvesting engines is pretty minimal. The creation of the larger models is where the consumption primarily comes in, and then the data centers that run them are servicing millions of inquiries a minute making the concentration of consumption at a single point significantly higher (plus they retrain the model there on current and user-fed data, including prompts, whereas your computer hosting ollama would not.)