I think you can replace “Europe” with anything here. Free Software is everyone’s best path to digital souvereignity. (Unless you’re big tech and have an infinite amount of money to develop things in-house.)

Nice. Now try PieFed and MBin 😊 They should have some more compatibility beyond their own niche.

And please broaden this beyond AI. The attention economy that comes with social media, and other forms of “tech-feudalism”, manipulation, targeting and tracking/surveillance aren’t healthy either, even if they don’t rely on AI and machine learning.

I’d say you’re probably right with the appointed internet sheriff. Quite some people here strive for a nice atmosphere, and Lemmy to be a particularly good realization of social media. That directly translates to the rules we make up for specific corners of it. And since social media is all about attention, this is the “handle”. I think it’s Privacyguide’s obligation. They can choose which place they call home and whom they bond with.

Seems this now had an effect: lemmy.one is shutting down in 90 days.

Nice idea. I’m not really sure if this is fast enough, though. The article doesn’t contain a lot of information about things like memory bandwith. But the 3.5 T/s for a 7b parameter model is about the same ballpark my 9 year old Skylake CPU does with DDR3(?) RAM (at double the energy consumption). It is a bit slower than I can read, plus there is a good amount of waiting for it to ingest the question / prompt.

As they write in the article… What’s a compelling app? In fact what even uncompelling software (copilot aside) makes use of this?

Ich frag mich immer wer das so bemerkenswert findet. Also das ist ja jetzt seine zweite oder dritte Aktion dieser Art?! Aber mich wundert es nicht besonders, dass Betrug funktioniert… Das tut er ja regelmäßig auch bei anderen Leuten.

Und ich find die ganze Urkundenfälschung, Medallien und so auch ein wenig fade. Warum nicht mal die eigenen Punkte in Flensburg entfernen lassen? Oder 'nem unsympathischen Politiker oder Großkonzern ans Bein pinkeln? Das wäre wenigstens unterhaltsam.

I think it’s kind of hard to quantify the “better” other than measure perplexity. with Q3_XS or Q2 it seems to be a step down. But you’d have to look closely at the numbers to compare 12b-q4 to 27b-q3xs. I’m currently on mobile so I can’t do that, but there are quite some tables buried somewhere in the llama.cpp discussions… However… I’m not sure if we have enough research on how perplexity measurements translate to “intelligence”. This might not be the same. Idk. But you’d probably need to test a few hundred times or do something like the LLM Arena to get a meaningful result on how the models compare across size and quantization. (And I heard Q2 isn’t worth it.)

Wow. Also stimmt nichts was in dem Text steht? Dann ist es ja explizit nicht “by Bank” sondern “by” externer Dienst dazwischen. Der gleichzeitig noch Vollzugriff auf mein Konto mit all meinen Geld bekommt? Also vertrauenswürdig finde ich das nicht. Und direkt von Konto zu Konto isses ja dann auch nicht. Der SWR behauptet ja da wär kein Dienstleister…

If you trust the measurements in perplexity which people occasionally publish on the llama.cpp project page or in some papers, the ability to run a larger model completely dwarfes the slight loss introduced by the quantization. That holds true to a level like Q4 or so. If you look at the perplexity numbers, they’re not far away from 8bit or full precision. And then it starts to deteriorate faster with Q3 and Q2, if I remember correctly. I always try to fit the largest model I can at something between Q4 and Q8. I can’t notice any major difference to full precision, though theoretically every step of quantization is going to lower quality by a tiny amount.

Und welche Bank bietet das an?

Denke ich auch. Ich hab noch einen von Primus und ein paar Werbegeschenke, mit ähnlichem Design, und ich würde sagen die halten alle ungefähr gleich lange warm.

Thanks. That is a good read.

https://help.zorin.com/docs/getting-started/upgrade-zorin-os/

I was recently made aware of Binoculars and it seemed to have a better than average accuracy. At least with the few texts I fed in to test it. But you’re right. Generally speaking and for most use-cases, they’re all next to useless.

I’m not sure if that’s LLMs. I long suspected people to copy popular Reddit posts/stories here to gain attention. I’ve searched for some and they don’t all seem to be dumped here from other places, so it has to be something else. But so far I didn’t use any LLM detector service to find out if people made them up on their own or used ChatGPT… We might want to do that to gain some more insight.

But it’s very annoying. I’ve unsubscribed from asklemmy and several other popular communities, because of this.

I’ve never had a look at the code of a major commercial platform, so I don’t really know whether those options do something. But sure. I’m not relying on it too much. I regularly decline and switch everything off. And I hope they at least honor my setting to not sell my data to third parties… But I don’t invest a lot of time into this, I don’t trust them so I rather avoid giving them too much data in the first place.

I’m curious how they’re going to monetize services like ChatGPT. Whether they put ad banners on the website and app, or make the chatbot recommend certain products to you…

Yes. I’m mainly looking for options for the charge controller and batteries part. (Preferably good, reliable and cost-effective.) I already found a solar panel, and I don’t necessarily need an inverter and 230V AC. I think something like 12V would be fine for some applications, if I got battery monitoring and over-discharge prevention.

I don’t think there is a valid reason to this. I mean they offer several other services as well. But I think many, many people use it instead of other web application firewalls, because Cloudflare are good snake-oil salesmen and everyone else is using it too, so it must be valid. And there is a free tier, and nobody cares giving away their and their customer’s private data and introducing one big single point of failure to the internet…