AppleMango
well if they're proper journalists then they shouldn't, journalism is not the job where you push your opinions into others.
And the fact that it hasn't been used before isn't what matters.. What matters is whether it actually has any real chance of causing disease. This just makes it seem like fear-mongering.
apparently audio and images are more efficient compared to text for multimodal models?
This is probably a case of heavy survivorship bias. Some don't get caught, and the ones who get caught doing actually crazy exploits we don't hear about.
Apparently it requires a Microsoft account to work, so local-only accounts are safe (in the case of patched windows ISOs). But this only disabled GDID and there's still no guarantee that there aren't other idenfiers.
The only part of "just works" that's still "just" working is that windows comes installed by default on the vast majority of laptops. I dread having to use windows each time I have to, not because linux is that much better, but because windows is really just that much worse.
and how do you think the social account was linked with said identifier?
arguably more stupid for using windows and assuming there was any level of privacy
Possibly, but how often do you use WiFi on a phone?
Basically whenever I'm at home or a relative's house?
fast ≠ widespread or consistent. mobile data tends to have excruciatingly low upload speeds (at least in my region)