

afaik brave has a feature that lets users upload their browsing data to brave for it to “improve its search engine”, maybe it’s that. in that case brave browser users are distributed scrapers.


afaik brave has a feature that lets users upload their browsing data to brave for it to “improve its search engine”, maybe it’s that. in that case brave browser users are distributed scrapers.


average people know next to nothing about the memory cartel


i personally enjoy my privacy rights. it’s not like police here don’t use for example dashcam footage. they do. so our privacy rights are already sometimes being ignored in favor of practicality. but glassholes go a step too far. i do not want to be surveilled by a private company 24/7. i do not want male-to-male transgender (a looksmaxxer) chad broski to record women without their consent and then post the video online as power signaling, and to goon over it. make it a free and democratic association using open source hardware and code, and open weights models hosted in the respective country with strong privacy guarantees, and I’m maybe on board again. roll it out very slowly, address issues as they come up and don’t make the product even worse for privacy before regulations can step in, only to call it inevitable and the future and thus regulation-proof when the law has time to catch up. legal repercussions should be severe, as you are violating a fundamental human right.


if violating privacy rights was so important then it should still not be done by facebook


smart glasses are not hammers. aside from the disability aid angle, they are primarily used by perverts for perverted use cases, and facebook wants to double down on that angle with face recognition, so they know their audience.
if these were prescribed by a doctor I would react very differently
I’m staunchly against normalizing being recorded all the time in public


it’s much more innocuous, to the point that many people can’t even tell you’re recording or wearing anything special that’s sending your image directly to facebook, especially if you drill out the notification light (which many perverts do do)
generally here in germany there’s a right to privacy if a person is your primary subject, even in public. you need their consent. there’s a difference between people being in the background and only incidentally and going up to people and “appreciating” them with a recording that facebook sees and also the whole world sees because you post it online as a way to signify your power over innocent women


blind people are actually decently assisted by smart glasses. it’s just a shame they’re mostly used by perverts. there should be a state-sanctioned version only for blind people, with limited features, and an open weights AI model hosted somewhere in the same country with data not used for training and such.


about the private security cameras, somehow amazon smart doorbells are still a thing though, and fairly common


an underrated feature in duckduckgo is easily toggling between international search results and local (my country) search results. both are things I want often, but I much more rarely want to select any other country, so it makes sense for the toggle to work the way it does. also noai duckduckgo is a blessing.


i haven’t had to switch to google in years. when I can’t find something it’s usually because I am looking for something from over a decade ago that nothing can find. I guess you have to search differently with alternative search engines, and I still have that proficiency.


✨ AI slop ✨


unless they eventually don’t need foreign currency anymore because their own country produces everything they could ever need or want


Podcasts have ads directly in the audio, similar to sponsorship ads in YouTube videos and on other platforms directly in the video the creator uploads. I was wondering when an open source solution to that would come out, similar to sponsorblock for YouTube, but I guess Spotify is first.
Podcasts are still slightly different, as the easy pliability of audio means podcast hosts inject ads based on your location and other factors into the audio stream instead of just going with one single global ad for everyone. But you still only get the audio stream from the regular podcast host, which is why classic ad blockers don’t work. A community-run solution might be "for podcast id 123 skip from audio fragment “sponsor” to “welcome back”. Just purely going by time (skip from 10:00 to 10:30) wouldn’t work, as ads can be longer or shorter. There can also be multiple ads back to back even if they’re a standard length.
A service like Spotify at least partially hosts its own podcasts and probably does the ads for those podcasts itself, so it can just leave them out. If Spotify proxies podcasts from elsewhere, it can detect (via querying the podcast multiple times from different locations) where the dynamic bits of the podcasts are and cut them out. That could also be a community solution, though it would require more computational resources. A centralized service like Spotify makes it easier. Spotify is also AI-happy, and what they’ve actually done might be to just train an AI on ad segments, and let the AI cut those segments out.
By the way, YouTube already skips some creator sponsorships with their premium service.


Private trackers or good (paid) Usenet indexers are two solutions for *arr tools specifically. If you’re manually adding files then just get them from trusted uploaders, manually check the upload before getting it.


kimi k3 is going to release on huggingface tomorrow. it will be open weight, assuming nothing gets between now and then. they just wanted a short exclusivity period.


people who just got into hiking and don’t have a good mentor don’t know that


these are hypervisor cracks, which previously required disabling most security features and running software with system privileges on windows. now on zen 4 and above on linux they just require a modified proton build, which you can even make yourself.


GLM 5.2 on openrouter costs $0.93/M input and $3/M output. claude opus 4.8 costs $5/M input and $25/M output. claude is only a “good deal” if you consider the subsidized plans for private users that let you use 10x or more the price you paid for.
That being said, I’m not a fan of AI at all.
for our cats what helps is CBD oil. it’s specifically meant for calming animals, and comes in various concentrations, so you can adjust them to an animal’s body weight. should have sensible concentrations and price/performance up to medium dogs, but of course light cats make it much easier to stretch a vial. it also comes in at the very least natural oil or salmon oil, depends on the animal which they like better. our cats don’t like the salmon very much.