IMHO LLM isn’t coherent with independence. That being said I wrote quite a bit on self-hosting LLMs. There are quite a few tools available, like ollama itself relying on llama.cpp that can both work locally and provide an API compatible replacement to cloud services. As you suggested though typically at home one doesn’t have the hardware, GPUs with 100+GB of VRAM, to run the state of the art. There is a middle ground though between full cloud, API key, closed source vs open source at home on low-end hardware : running STOA open models on cloud. It can be done on any cloud but it’s much easier to start with dedicated hardware and tooling, for that HuggingFace is great but there are multiples.

TL;DR: closed cloud -> models on clouds -> self-hosted provide a better path to independence, including training.

How is asking to justify a position trolling? You are the one who claimed that Danish law is better than GDPR. I didn’t claim you lie or that law elsewhere was better, I solely asked for the proof. It’s not because I mistrust you, I just want to learn and you saying it is so without an actual comparison is not enough. If you don’t want to help that’s perfectly OK you can just say so. It’s fine to say you prefer Danish product because they are better and refuse to give proof that it’s the case. It won’t help me nor others though.

It’s the Privacy community on Lemmy, I bet others would love to learn too.

Bought a 2nd-hand Pixel 8 to put GrapheneOS on it, not sure if that counts. Feels old, more ecological, cheaper and more private. Not sure how repairable it is but in theory I can use it for up 7 years so hopefully by the time I need to repair it I wouldn’t even want to.

Thanks, skimmed through https://www.recordinglaw.com/world-laws/world-data-privacy-laws/denmark-data-privacy-laws/ but it’s quite difficult to do a “diff” between one and the other. From reading it I didn’t notice significantly better for my normal usage but I’m not a lawyer. It also makes me wonder, if you have done it, how do you know it’s not better than say another random EU country also national specific modifications, e.g. Slovenia? Is there any “benchmark” somewhere that identifies which national changes are better?

Why would anyone who care about privacy not provide a new alternative WITH a benchmark comparing the newcomer to what already exists (e.g. Google Android, GrapheneOS, PostMarketOS, /e/OS, iOS, etc) and 3rd party audit is beyond me.

Which Danish laws go beyond GDPR?

CoMaps can export a GPS Exchange Format (GPX) route. See https://wiki.openstreetmap.org/wiki/GPX and that can be used to get most of the data your need. I imagine you could write a spreadsheet or online to do your visualization. I also bet others already did it.

I’ve done something like that for hikes to plot the path in 3D for WebXR.

As others points out GadgetBridge if you want additional data, typically biosensors for heart rate monitoring, is great to see which hardware is supported.

just that the mobile Proton Mail app does not support fulltext search. I know why, but I still think it’s doable the same way as in the web browser

If your mobile has a modern Web browser I’m pretty sure you can do full text search in there too.

Also FWIW it’s a constant struggle for everyone.

Corporations do their very best, both technically but also with marketing and lobbying, to make it nearly impossible. We have to learn, help each other, vote and it will never stop. Still, each step mattes so kudos on even attempting.

I’m not sure for what matters it makes a difference, namely https://en.wikipedia.org/wiki/Principality_of_Sealand#Legal_status

Also AFAICT since https://en.wikipedia.org/wiki/HavenCo my understanding is that P2P and distributed technical solutions (torrent, Onion routing, I2P, bitcoin, etc) improved enough that it wasn’t really worth it.

Makes me curious, I assume within the EU with GDPR it would be roughly equivalent.

What’s the difference between EU countries then and why?

“Fear, uncertainty, and doubt (FUD) is a manipulative propaganda tactic used in technology sales, marketing, public relations, …” https://en.wikipedia.org/wiki/Fear,_uncertainty,_and_doubt

Correct I’ve done so few times on my PeerTube instance using OBS on desktop and another client on mobile.

Sadly FUD as ANYTHING that is NOT increasing profit for surveillance capitalism, i.e Google, Meta, etc is a win for privacy!

Of course /e/OS could be better, GrapheneOS could also be better (including on security) but the big picture is that still ANY of those solutions is making surveillance capitalism, the loss of privacy for profit and power, less efficient. That’s good for all of us who, being on Lemmy or other federated instance, believe we do benefit from having more privacy, or at least not trading it away.

TL;DR: be inclusive, bring others up, don’t be exclusive aiming for perfection none of us can attain.

Love that quote “We’re going to search your house. We don’t have a warrant, but we paid your landlord $100 to give us a spare key. So now we’re searching your house without a warrant,” by Laperruque

Isn’t it “source: model” basically roulette? We go back to the initial problem. Also anything else that is not model might also be hallucinated if at any point the string that gives back “source:” goes through the model.

Right but if one person keeps on giving me wrong answers, knowingly or not, my distrust in them in not linear. They’ll have to “earn” it back and it’s going to be very challenging. If they do learn though, then it might come back faster. In this setup I have no guarantee of any progress. There no “one” in there trying to fix any mistake.

Can’t it source other LLM outputs as “verified source” and thus still say whatever sounds good, like any LLM? Providing “technical” verification, e.g. SHA, gives no insurance about the content itself being from a reputable source. I don’t think adding confidence and sourcing changes anything, the user STILL has to verify that whatever is provided is coherent and a third party is actually a good source. Thanks for making the process public though, doing better than OpenAI does.

CoMaps / OpenStreetMap / local public service for public transport

So glad I moved from iOS to /e/OS (last year) then to GrapheneOS (last month).