You’d think you’d want it recommending porn in that case to use later as blackmail.

And yet the “free” market can’t be trusted to correct itself and fill the gaps if we allow a giant megacorp to fall.

A Microsoft-oriented news outlet.

Think similar to MacRumors/9to5Mac/AppleInsider for Apple.

I think it probably doesn’t matter what he wants, it only matters that the data exists at all. If the owner is not giving permission, that’s one thing. But I’m inclined to believe that those American 3-letter agencies aren’t the sort to ask permission.

All it takes is one disgruntled systems engineer who thinks they don’t get paid enough. An agency comes knocking with a sizable offer of cash, and they’ll get the backdoor they want.

E2E encrypted messages in Matrix contain more user metadata than alternatives like SimpleX, nothing scary but a MitM is able to see origin points, destinations, and times of messages. Server to server, if you’re using E2E encryption, it relies on trust that the other server is not compromised.

And it seems Matrix.org is not the best at security disclosures: https://soatok.blog/2026/02/17/cryptographic-issues-in-matrixs-rust-library-vodozemac/#matrix-response

Nothing federated is private, mind. Even with E2EE on in private rooms for specific messages, Matrix still relies on a constant information feed during use that can be used to deduce who is messaging whom and when, even if the content of the message itself is encrypted.

This is why it’s important to seed your online activity with deliberate false identifiers. That way, no one with bad intentions will learn that I work for the ICE office in Santa Fe, and always attend church every Sunday (when football isn’t on, of course).

I can’t be arsed to keep up with changing the config of the browser I didn’t choose every time the device updates with new admin-defined settings. That’s all.

On my personal device I still use Google right now for consistency/because change is hard, but I set the default behavior in Firefox to exclude AI results.

I’d just prefer to use/support a search engine that abstains from AI entirely, regardless of whether or not you can turn it off. I don’t want to be a happy customer of companies that still try to weasel that stuff in, because they won’t stop at a toggle. They never do.

Yeah but you can do the same with Google as well (just can’t be arsed to change the search settings on my work PC), it’s the principle of the thing.

That’d be my guess.

I have never once went out of my way to use Gemini, for example, but having it appear in every other Google search with some lying bullshit to spread is probably driving up their engagement numbers.

I really should stop using Google. But can’t use DDG either because they just use Bing and their own AI service as well, and so does Kagi.