phyzome

Maybe! I expect a lot of animals construct narratives to make sense out of the world around them.

What's great about this is that the primary thing exposed (credit card info) is also one of the least worrying things to have exposed, as a user.

You can just get a replacement credit card.

Since we know that essentially nobody is using homomorphic encryption, the original point stands.

Boy, people sure do love turning data points into stories.

So, completely ignoring the abusive CEO as an issue of its own, I think there's one very important aspect of AI usage that's already clear from the article:

« Workers describe an environment where employees have genuine concerns about security or being asked to develop security products with AI, but feel unable to voice them »

It's very hard to build reliable and secure products in a fearful environment. This sort of thing has come up in root cause analyses of disasters. We don't need to know whether Proton has defined "guardrails" for AI usage, or what the processes are like; there will be problems, gaps, and employees will be unable to raise those issues with management. That's all I need to know, really.

You can absolutely shake people out of their ways without being a raging asshole.

For the kind of end-to-end encrypted support they advertise? Tuta is the only one I know of.

Email more broadly? I can heartily recommend Fastmail.

I don't really send anything sensitive over email if I can help it. Signal is a much better choice.

Posting the CEO's own words seems pretty fair to the CEO.

...no? That's just called being a crumple zone.

There have been a number of signs over the past few years that they are very much not the good guys.

Take a(nother) look at the post he linked to. Comments were broken previously, but they're showing up now.

Pyon indeed seems like a complete asshat.

No, HN has a braindead feature that autostrips certain words from titles.

New technology bears additional focus.