Cursor is just an IDE (integrated development environment), you can set it up to use all sorts of LLMs either directly through Cursor, or with your own API keys for the sources.

This file content just goes into the initial context to help the LLM act how you want.

I am not sure why anyone would use an AI code editor if they aren’t planning on vibe coding.

Vibe coding means only looking at the results of running a program generated by an agentic LLM tool, not the program itself - and it often doesn’t work well even with current state-of-the-art models (because once the program no longer fits in the context size of the LLM, the tools often struggle).

But the more common way to use these tools is to solve smaller tasks than building the whole program, and having a human in the loop to review that the code makes sense (and fix any problems with the AI generated code).

I’d say it is probably far more likely they are using it in that more common way.

That said, I certainly agree with you that some of Proton’s practices are not privacy friendly. For example, I know that for their mail product, if you sign up with them, they scan all emails to see if they look like email verification emails, and block your account unless you link it to another non throw-away email. The CEO and company social media accounts also heaped praise on Trump (although they tried to walk that back and say it was a ‘misunderstanding’ later).

Some people like it for the ui and they recently announced the ability to turn off all ai features.

Unfortunately so is Visual Studio and VS Code, yet we don’t say anything made with them is ‘vibe coded’. The text, big and bold, right the top of the screen for VS Code is literally:

But what does that have to do with consumer privacy?

I see their repo is open source. Is there any actual evidence that the sketchy machine generated any part of it?

Not an issue, per se. In order to keep the team small they built most the app in a single codebase. It’s mostly web code, and the apps are wrappers for it. So it keeps it unified between all clients but it definitely feels like a web wrapper, so it can feel a bit slow or clunky.

I have tried Tuta out, it’s fine from my very limited use, but kinda locked in in ways I don’t really care to pay for. Last time I saw it brought up some other folks were recommending mailbox.org. I don’t know about it too much, but might be worth looking into as well.

@sunzu2 it feels janky as hell, it’s missing advanced features (someone in the other thread asked about Sieve filters), and it doesn’t support non-Tuta clients. their development cycle is so slow I can’t count on any of these features cropping up anytime soon.

with those criticisms in mind, Tuta’s still approximately the only credible choice remaining for threat models where end-to-end encryption is important. we desperately need better fully open source options for this.