

Hey free GDPR fine!


Hey free GDPR fine!


“your rights don’t apply because you’ve been mean”? What are these grounds?


Do you consider AI good enough to substitute human interaction?
I don’t think it’s a matter of capability. Even as a neurodivergent introvert, I recognize in human connection one of the most fundamental aspects of our life on this earth. You may get answers that a person also would’ve given you, but think of all the people you won’t get to know, groups you’ll never be a part of.
it surprised me how complete and articulate answers are.
That’s their greatest trap. They always need to sound confident to have the best chance of providing a good answer, so that when they don’t know or you get an unlucky seed they just lie/pretend very convincingly. Never rely on LLM output as the sole source, even, especially, when it sounds plausible.
It is very important to remember that even if they “talk” like people, they are not people. Using them as a search engine on steroids is alright, but if you ever find yourself asking “what do you think about…” you’re using them wrong. Never treat them as a friend, ask for advice, talk about your emotions, etc. This is easier said than done as a language-machine activates out brain in all kinds of funny ways; forgetting you’re talking with an algorithm is very very easy.
Also, don’t listen to people who say you’ll be “left behind” if you don’t jump on the train. That train is not headed to a nice place.
All that said, they are sometimes marginally useful. I would say that cloud AI usage is never ok and self-hosted models can be acceptable, depending on their usage. Look into Ollama and Unsloth. You don’t need that much RAM/VRAM to get something workable, what are your hardware specs? I can suggest some models


I have also been thinking about this, because as far as I know you’re correct - at least under US copyright law, AI output is not copyrightable, and since you don’t own it you can’t license it. All AI code is public domain (again, at least in the US). It’s surprising to me that there has been little discussion about this, and Linus himself has not addressed it


90€/mo to self-host something? Lol


That is indeed work, yeah


I am not sure it was initially created for evil, I do believe that its discovery was mostly academic and in that it was, and still is, a pretty significant/interesting one. It was very quickly and very strongly made into the perfect weapon for the information age for sure, but i doubt that the original researchers wanted that, or just really thought of the implications outside of information theory. But it was indeed Google, surveillance capitalism incarnate, that initially gifted transformers to the world and kickstarted the whole thing…
Edit: I realised that I inadvertently made this comment about “AI” by way of pronouns when I’m really talking about LLMs. Google of course did not invent AI, and I more strongly disagree that AI in its actual sense was invented for or is inherently evil.


Depends on the workload. This probably would not be great for gaming or audio production, but it might be for video editing and local LLMs


There’s a thing called search engines, it takes about 20 seconds, much less than waiting for a reply you asked for in a mocking tone. What about “security analysis tor protocol”?


So it’s closer to an ML shader than actual supersampling?


Wait, what is even the point of using an upscaling technology if it makes your game run slower than without it? Why wouldn’t I just run native res at that point? Is their only point that “it looks better” in their minds?


noblogs.org is still reachable, for now


Whether a system(d) service is activated by default after you install a package is a distro-specific choice. I know Debian and derivatives do, so it will be started at the next reboot; while Arch and derivatives don’t, requiring an explicit systemctl enable [--now] something.service to run it. The devs of the distro try to pick what they think their userbase wants and after the initial decision is taken it’s not a behaviour that can be changed lightly.
Regarding running as root, sometimes it’s just the default configuration that can be changed to a dedicated user, other times it’s required by the privileges they use (e.g. if they need to open a port below 1024)


But while the AI was ready to give up several times, it did keep adding debug code and analyzing it faithfully when I pushed.
Well, no shit, all they can do is follow instructions. We just really can’t help avoid anthropomorphizing something that speaks language and think it was a good boy for stopping whining and trying again


It’s not mentioned because it’s not required, yubikeys in general mostly leverage pre-existing “smartcard” facilities


Huh, for some reason I assumed XMPP didn’t support video chat. Does it also support screen sharing or do you need a fake input and send it as regular webcam video?


I don’t see language being passed in the example code, so it may be agnostic?


Mandriva/Mandrake might be older than you…


Anecdotally, I’m seeing a lot of bot traffic masquerading with a macOS user agent
Probably a combination of that and general human incompetence in using or developing the tool