Unusable 3151 ⁂

Snagged it!

There are plenty of outlets that specifically do not use AI in their writing. Those are what I fill my feed reader with.

It’s not like I’m going to boycott them, but it’s important to always push because they CAN do better. Any “risk” that they are not handling, they are offloading to the workers themselves. It doesn’t just disappear. I’m not saying they should be only hiring full-time devs at competitive silicon valley salaries, but part time on a living wage with worker protections is an important bar to clear for all entities that seek paid labor.

Yeah, this is such an anti-labor thing for them to do. I hate seeing stuff like this in general, but from open-source projects it hurts way more.

I totally forgot about “Skannerz”!

Thanks for rushing back a bunch of my childhood memories

Because of who decides how such verification gets built. Just look at what is happening with Android apps to know how big tech companies will handle it. https://f-droid.org/2025/09/29/google-developer-registration-decree.html

Whatever Apple, Microsoft, and Google decide it does.

This would require “verified” operating systems: No alternative Android-based OSes. No Linux phone. Hell, no Linux desktop, or at least we would be stuck with a big-tech-built proprietary web browser.

This is a terrible idea.

barebones, but you can use RTP with pipewire

The one OP mentioned is the 2nd-gen model that came out in 2019

Yes.

First, use closed-loop systems whenever possible, so they can act independently and no data is ever collected.

Second, when closed-loop systems are not enough to provide the desired effect, only use sensors that make it physically impossible to collect data that can be disambiguated. Nothing should ever be deployed that can be used to identify an individual with a firmware or software update.

Third, if it can’t be done in either of those ways, don’t do it. There are certain conveniences that are only possible with surveillance, but they are just that, conveniences, and I think that the vast majority of “smart” things can be done in the top two ways.

I’m fairly certain that there is an issue with AMD and HDMI licensing that is a barrier here

Stop hoping and blaming and get in the fucking trenches.

Yeah, F-Droid works. I decided to go with Obtainium. The GPS signal is good enough in my city. The maps app itself is serviceable. It’s essentially a UI for OSM data.

I have the same phone. It’s fine. I keep my Pixel phone with GrapheneOS on it at home for banking apps, Briar, and a couple other things. A few things to note:

1. The slider on the side doesn’t turn off the GPS antenna, so you can still navigate with the GSM module unpowered using the default maps application or another offline-capable app, which is nice.
2. The company advertises the slider as a tool to protect you from “harmful 4G waves”, which is a bit concerning for their direction as a company.
3. Mudita decided to completely abandon releasing their software under an open-source license for this phone, which sucks.
4. InkOS is an open-source launcher that was designed for the phone and it’s great.
5. Software features that were advertised even before the launch of the phone are still not available not planned to be available (certain stock launcher customizations in particular)
6. Not the phone’s fault, but you can’t have Signal installed on multiple phones for the same account, so I have to turn on the slider on my phone to make a Signal call at home instead of being able to use my WiFi-only Pixel, which is a bit annoying from a privacy perspective. EDIT: I can also use my computer as a workaround, it’s just a bit of a clunky solution.

As a John in both name and in relation to my main project at work, I am working my ass off to write comprehensive documentation and to train the rest of the team to take over. I don’t wanna be working on this same shit in 5 years.

I also have a OnePlus 6 with Mobile NixOS. I haven’t been able to get audio or camera to function, so it’s just a toy on m desk at the moment. Other than that and a few UI quirks, it’s serviceable.

It’s important to make a distinction between the definition of “open source AI” canonized by the OSI that doesn’t require open training data, and models where all of the training data used is also made available.

Separately, the tools most people think about when they hear “AI”, generalized generative AI models, only exist as capitalist surplus, and we shouldn’t be defending them. Hyper focused AI tools such as the Te Hiku Media project to create speech recognition tools for the te reo Māori language are unequivocally good, and we should be making a lot more projects like this.

mineral aerosols.

here is a recent paper trying to figure out the composition from ice core samples for a given age range and location

https://pubmed.ncbi.nlm.nih.gov/37061058/

You’re correct! It is Cherenkov Radiation; specifically from the muon (or electron or tau) that is a result of the neutrino interacting with the nucleus of an atom because Cherenkov Radiation happens with charged particles.