Tangent5280

joined 2 years ago
[–] Tangent5280@lemmy.world 1 points 3 hours ago

Oh shit, I thought it was OP just being a dummy. Turns out I'm the dummy

[–] Tangent5280@lemmy.world 2 points 11 hours ago

Yeah, because people want their SO also to know how to use the laptop when they need to.

[–] Tangent5280@lemmy.world 1 points 12 hours ago

That's called a market opportunity.

[–] Tangent5280@lemmy.world 8 points 3 days ago (6 children)

Hey that's a pretty good idea. I'm stealing that.

[–] Tangent5280@lemmy.world 5 points 1 week ago (1 children)

Which poor guy has half a Dingdong?

[–] Tangent5280@lemmy.world 6 points 1 week ago (3 children)

I was so excited to get my grubby hands on that music only to later learn it was hundreds of terabytes...

[–] Tangent5280@lemmy.world 1 points 1 week ago

What exactly is this thing, I'm curious now

[–] Tangent5280@lemmy.world 1 points 1 week ago

The ensuing explosion took out two thirds of New Jersey

[–] Tangent5280@lemmy.world 1 points 1 week ago

Bosenana AudioBananica DolBananAtmos Bananaheiser

[–] Tangent5280@lemmy.world 19 points 1 week ago

There's always music in the banana chain... or something like that

[–] Tangent5280@lemmy.world 3 points 1 week ago

I meant no insult, just that the comment had sentence structure that I've seen in LLM responses.

To be honest, I kind of prefer this, because the alternative on many computer centric discussions is just people assuming I know everything they know and just throwing jargon left and right. Atleast here I can infer details from context.

[–] Tangent5280@lemmy.world 1 points 1 week ago (1 children)

Is it on F-Droid? It's a PITA to install play store apps because I'd have to install play store (bleh) so I'm hoping symfonium has a github or something so I could use Obtainium.

 

People are noticing that their phones are getting an app called "Android System Safetycore" auto-installed without notice or consent. Check your phone for the same, it is likely it's a slow rollout instead of every device getting it installed all at the same time.

Google has all the same old reasons that they drone on about, but the actual reason is likely to harvest your messages data for training AI models.

Uninstalling seems to remove the application, and there aren't any malicious activity reported so far as I can see, but naturally that can change anytime.

Has anyone noticed this in their applications lists? Did straight up uninstalling them work? I've had some trouble removing systems apps in the past, but uninstalling this one seems to have worked straightaway - I don't see them in the list anymore.

URLs below for Reddit posts about the same: From 2 months ago: https://old.reddit.com/r/antivirus/comments/1gpdhwz/guys_help_some_app_called_android_system/

From 2 days ago: https://old.reddit.com/r/privacy/comments/1idjbdi/googles_new_app_will_help_warn_you_about_nude/

 
 

Basically title. I waited on installing F droid for a long time because my phone threw many scary warnings when I tried a long time ago. But now I have it, and I got some fossify apps, but since there is no "Editor's Picks" on F- droid I dont really know where to go from here.

What apps do you recommend I install first to remove my dependence on closed ecosystems?

What is my vulnerability surface ie, which sort of apps should I watch out for?

Are there any bad faith companies in the open source sphere?

 

Does anyone here use any Open-Source Workout Trackers? I've been using hevy, but their high fees, the fact that they are a company that holds my health data and has made no commitments to open source, User privacy, or fair trade practices like user data import/export has me looking around. I wanted to see if anyone had reliable open source alternatives.

Tell me your workout tracking stories here! Tell me what you liked and what you disliked.

 

Is using Voyager giving Chrome an opportunity to harvest user data? I'll take whatever you know about the Voyager dependence on chrome.

 

Useful because now you'll be able to tell that something is human-generated instead of AI-generated, and content creators and people with a large public presence will now be able to police their own likeness being used by randos.

Scary both because now whistleblowers or reporters could get their cover blown because the image has metadata linking them to it; or they could strip off this metadata and get the evidence dismissed entirely as fraudulent; and also because of the possibility that any regulatory government body that enforces C2PA will also determine what is real and what is not, meaning anyone on the inside will be able to generate AI content and pass it off as real to the vast majority of the population.

Can't help but think they shortened it to C2PA instead of CCPA because of the similarity in acronyms of the latter and the big bad no-privacy country.

What do you think? Non-issue, Slightly concerning, or apocalyptic?

view more: next ›