The right way to use "AI" is to not use it.
this post was submitted on 03 Oct 2026
-11 points (29.6% liked)
Privacy
51573 readers
336 users here now
A place to discuss privacy and freedom in the digital world.
Privacy has become a very important issue in modern society, with companies and governments constantly abusing their power, more and more people are waking up to the importance of digital privacy.
In this community everyone is welcome to post links and discuss topics related to privacy.
Some Rules
- Posting a link to a website containing tracking isn't great, if contents of the website are behind a paywall maybe copy them into the post
- Don't promote proprietary software
- Try to keep things on topic
- If you have a question, please try searching for previous discussions, maybe it has already been answered
- Reposts are fine, but should have at least a couple of weeks in between so that the post can reach a new audience
- Be nice :)
Related communities
much thanks to @gary_host_laptop for the logo design :)
founded 7 years ago
MODERATORS
Maybe self-host a model?
Checkout Lumo from Proton. Its quite good and private. You can read about how it works on their blog. Otherwise selfhost it.
Grab llama.cpp. Self-host on your machine in a VM that has GPU permission but not network. Presto, no big-tech data collection. No fingerprinting. No profiling. No ads.
I've used it like that for language translation. Sometimes I wanna read a German, Japanese or w/e page. Or say a few words to someone in their own language. Which my human brain can't do. Local models do a more than acceptable job for me.
You have to be really, really super careful about confabulation, sycophancy & other probs. Do not get into the habit of asking a q to the model and believing it. And don't use it agentically. But for page translations? Other low-importance tasks? Useful.
changes the fingerprint each session going to help
It's gonna be super hard to totaly defeat fingerprinting. And you'll never know if you really have. Local model + deny network, you can be almost totally confident. "Almost", b/c nothing is 100% perfect, ever. But you go from almost 100% chance of surveilence, to almost 0% chance of surveilence.
Forgot to say, this is better wth a GPU. You CAN use CPU + RAM, but very slow. Even a 6 or 8GB GPU is OK tho for a small model. The bigger your GPU the better. But a mid range gamer GPU is adequate for many practical purposes like language translation.
I guess this is part of the problem. Hardware costs are ridiculous a.t.m. i'll have to see what i can find on ebay.