I think that the point is it's entirely pointless building something like this into the email system. It should be a separate system that you can choose to use if you want it. Building it in just opens questions about exactly what they're doing with your data, despite their assurances.

For me the issue here is, why put so much time and energy into basically rebranding an LLM. I've seen LLMs running on RPi and android phones. Why not write a blog post showing how to run LLMs locally with existing tools for the best privacy instead and put more focus on their existing services. It just seems like they're jumping on the AI bandwagon and charging a premium for an already freely available LLM.

I see some benefits of AI like quality tts when using OSM and stt when transcribing/translating audio but other things like Googles AI answers and Microsofts Copilot leave me scratching my head wondering why consumer would want this

This is one of those situations. People would eat sweet food if they didn't know it's bad for them. In fact they still eat shitloads while they do know it's bad. Same goes for opiates, and there's a reason why they are regulated in most of the world (the usa of course is fucked up with this). So yes, let's boil the fucking oceans so dumb fucks don't have to learn how to write a fucking email. Let's say people have to learn what kinds of shit are included with "ai", they have to learn how much it actually costs, what all of the nft cunts are now doing, and then say if they want something or not. I mean, this is all just fucking gaslighting by now, right? I don't care what a fucking ignorant shit thinks about mostly anything - but this is a shitty company telling us 'you want it so hard baby' so it can pump that ai investor hose SO HARD