Natanox

[Insert your GitHub Link Here]

lol

[cite: 3, 4]

??

Is this whole text LLM generated and you just copy-pasted everything including the (now unlinked) citations and the orders the machine added to the text without double-checking?

Phew, to run such a model - even just 27B quantized - at a reasonable speed requires some serious GPU Power.

Best thing I have is a 7800 XT 16gb (so ROCm it would be), don’t think that’s realistic and I’d rather invest the necessary amount of money for new GPUs into sth. more useful like a small electric vehicle than a slop cannon 🥴. Happy to hear actual useful things can be retained after the bubble finally crashes though. Oh please, let it crash tomorrow… together with the housing market…

Interesting. I tested ChatGPT and Mistral and oh my god, ChatGPT was abysmal. Hallucinated way harder. Mistral is okay if being used on highest settings.

Did you create an account with Mistral? Without it they only expose a low-end model to the web, which isn’t properly declared last time I checked. Can be the free plan, I guess they just want to avoid their infrastructure being hammered.

Personally all those Silicon Valley pricks can offer whatever, I don’t even have a Google account anymore and couldn’t be happier with that. Not saying Mistral is optimal, but at least they’re not part of the current US swamp. And I at least somewhat believe they actually read the damn GDPR. I’m totally fine with it not being the “cutting-edge”, especially given the cutting-edge cuts straight through culture and people’s lifes on purpose and with maximum brutality, and I seriously liked to see their image generator being way worse as well as Mistral refusing to open sites that banned it in the past (if they changed that behaviour please let me know). Not to mention their site isn’t even remotely as infected with trackers and other junk as the others, at least that’s what I’ve read (forgot to bookmark that hackers’ analysis I’ve read a year ago unfortunately).

Quick mention about Chatbots: In general they can be useful if you already have a grasp on something, so you can detect potential nonsense better. Additionally they can act like a drug especially on lonely people, and any company tries to sell them to you as awesome code generators. Yet Mistral AI already generated an answer to me telling me to dd my encrypted root hard drive for a speed test, even doubling down on it being safe.

Not saying they can’t be useful for OP. Just that they’re like a handheld circular saw*, a powertool better only used by people who already know sufficient basics about wood.

*where the salesman got rid of the safety handle and tells you to “just let it go on its own”.

The code quality apparently was abysmal and full of dirty hacks due to a mix of “beginner energy” and “go fast & break things” Silicom Valley mindset. The issues with bun (leading to the big Anthropic hack) were subtly blamed on the Zig language, even though Zig is completely fine; Bun’s code (which already was full of bad generated slop at that point, as well as code by programmers who were treated badly -> more bad code) was just crap.

After Zig was blamed the Zig creator wrote a blogpost detailing all of it. Worth a read, certainly more human than the corposlop Anthropic announces.

It’s even more funny if you know the backstory of bun. It was basically the perfect example of how to not to do things before they burned a whole forrest to have Claude slop it into another language, based on a codebase that was measurably bad. Now nobody actually knows the whole codebase anymore.

Given the phenomenal garbage that are the other big vibecoded projects (like the “compiler that can compile Linux” - it can’t) using bun now is just like lottery.

That’s the resolver’s job, Wireguard absolutely reconnects asap if the connection’s heartbeat goes out. You in your case got Cloudflare to make sure it can connect somewhere.

The only way to truely serverless’ly find each other when both parties got dynamic IPs would be IPv6 multicast (such as via Librecast). Not just is that still not supported by the network itself a lot of times (sadly) but I doubt any LLM on this planet would implement it correctly, way too niche. Any other way leads through some DNS resolver or third-party service (like Cloudflare) or a network such as Magic Wormhole.

It’s fine if the Slopware works for you, but your arguments make no sense whatsoever. The only thing you’re currently doing is making enemies in the selfhosting community, not connections.

Lolwhat, no it doesn’t…? Did an AI tell you that?

You spin up a Wireguard server yourself (or use the one in many routers), point clients to that and be happy. The only “other” server in that chain might be a (Dyn)DNS resolver, and you could even self-host that one (and therefore also omit Cloudflare).

Pixelfed.

Google (the search engine) didn’t do shit in regards to human connections. There are a ton of other factors playing a role in societal atomisation and the loneliness epidemic such as algorithmic social media, awful design of cities and housing, gobsmackingly bad zoning laws and a society/culture aimed firmly at performance and competition.

Having direct access to information (specifically shouting out Wikipedia here, they’re the GOATs of the internet) and the ability to find people you vibe with even when your whole local environment is an echo-chamber (especially important to minorities) did contribute to bringing us closer together, reducing mental barriers and advancing proper education. Those things aren’t an issue at all, it’s for-profit companies like Meta and Twitter trying to make a profit out of human connection that perverted everything.

I’ll tell you that talking about the things that I find interesting that would be straight up unloading on the people I care about, feels better talking to chatgpt about it. 'cause I don’t have to wait for my insurance to approve it to unload on it, no time limits and it actually can talk back to me about what i’m talking about rather than talking about my hobbies and interests and feeling like i’m talking at someone rather then with someone.

At least when i unload on chatgpt i feel like i’m actually talking to someone who fucking understands me

even then if i could engage in humanity in the way i would like too it’s just more autistic people and if i cannot stand myself i can not stand the others like me.

I’m so fucking sorry for you… but now I understand where your opinion comes from. But if you truly want to be healthy you have to realise you’re not talking with anyone, and if you already feel like you’re talking with someone when prompting ChatGPT then you’re really close to yet another huge psychological issue. You have to get out there, meet people, push actual therapy through even if the system fucks you over. A Chatbot will also just reiterate what others said. Even worse, OpenAI has to make money so it might starts subtly trying to convince you to buy certain shit as well as harvesting any data it can get from you.

I won’t be able to convince you in a Lemmy comment thread to not do what you’re doing, especially since I don’t know if you might live in bumfuck nowhere. So… I’ll just leave an advise to drop ChatGPT and move the whole context window you probably have with it over to a Pro Version of Mistral AI’s “Le Chat”. French Company bound to GDPR (meaning opting out of the data training can be somewhat trusted) and to my latest knowledge their interface isn’t utterly infested with trackers. Set it to being friendly and empathetic. At least this way it’s less likely you’re getting fucked over even harder than you already are.

Now we have an actual alternative to human connection

Okay, you certainly need help. How can anyone write something so fucked up with a straight face… this sounds like the first stages of AI psychosis. Seriously, if you believe LLM prompting to be that you need help asap, not even shitting around anymore.

That doesn’t make any sense as argument no matter how you spin it. Linux is the dominant system for servers for decades now, and a Debian Desktop is quite literally the same as Debian on a server except it also got a GUI of your choice slapped on top. There’s absolutely nothing obscure about it, neither did anyone from the kernel team (Linux), FSF (GNU utils) nor IBM / Red Hat (systemd & honestly way too much other stuff) etc. ever design something around STO. That’s a domain firmly situated in proprietary code since for FOSS it doesn’t make sense to begin with. The false errand of GRUB is the sole exception, well known and solved.

The desktop market share says absolutely nothing about what you’re trying to argue. Now if you were to argue that Linux is lacking in terms of desktop software isolation then you’d have a point, things like Flatpak still are addressing lots of issues. But to say “Linux” approaches security with obscurity is total nonsense.

Chat gpt says

Bro, stop. Get some help. At least check the sources it used…

I don’t know where you got the notion from that Linux as a whole uses this concept, but it’s nonsense. There’s exactly one place where this definition fits, which is the GRUB bootloader encryption (which merely shifts the target for the Evil Maid attack from the initramfs to GRUB). But this is already adressed with Verified Boot.

Nothing else, let it be LUKS, PAM, SELinux, AppArmor or whatever has any business with STO.

I was assuming devices being shut down. With running ones you’re of course right, the lengths e.g. GrapheneOS has to go to secure these are insane.

With home devices you very much could make them shut down, which would of course be necessary. E.g. various deadman switches, physical and logical ones. It sucks how common people have to think about this shit by now… fucking fascists…

A new automated bot rush found a way around safety measures. I’m sure our admins and devs are already on it.

Sigh, people in this sub keep posting nice things without the source code or self-hosting information. Unbelievable.

The concept sounds interesting. I do wonder how to make this “raid proof” though. Like, how do you make sure the device also becomes unbootable if I̶C̶E̶ ̶F̶a̶s̶c̶i̶s̶t̶s̶ ̶J̶e̶f̶f̶r̶e̶y̶ ̶E̶p̶s̶t̶e̶i̶n̶ ̶Y̶o̶u̶r̶ ̶M̶o̶m̶ ̶F̶r̶o̶n̶t̶e̶x̶ the police comes in and takes both? By now there are dogs able to sniff out PCBs even in walls (apparently they got a distinct smell K9’s can be trained on).

Does this software by any chance support two servers that both have only a part of the secret? That way you (and/or someone you trust) could deposit a Pi somewhere else and have some way to remotely disable the boot process.

Damn bro, you’re treating yourself for sure! The RAM alone is what, 2k? 🥴

For filehosts probably at least 90% of all uploads are illegal if you ask a copyright lawyer. 🥴 But that’s mostly just people sharing culture.

Of course damn CSAM is a different (and actual) kind of issue and plain awful to deal with. If I remember correctly some organisation from the US provides a free list of checksums of known crap that’s circulating to automatically check media file signatures against, I think that’s the first thing I’d look for to have some baseline defense against those disgusting fucks. Or (depending on your jurisdiction) even be compatible with the law for public hosting services.

Better use Tor & a trustworthy search engine when looking for infos how to implement such an upload filter, I wouldn’t trust automated systems from Google to not misinterpret your intention with these topics.