Training LLMs is predicated on sucking up as much data as possible. Because the established chatbots have already sucked up all of the literary content on the internet the only new source of data is the users directly.

This will inevitably create a problem for privacy as making the LLM better will necessitate the intake of more and more user data. If it comes to a point where Lumo must compromise user privacy in order to improve it, I would rather not have it at all.

LLMs make people dumber by outsourcing critical thinking skills. They use literal tons of drinking water and the assocated datacenters are often established in places that are water insecure.

The reshaping of the whole US economy to cater to LLMs have irreperablely damaged the consumer technology pipeline and ensured the pipeline of the newer IT people have a harder time finding work and be overall less experienced handling simple issues themselves as more and more of the workload is shoved off onto chatbots.

LLMs are an overvalued autocorrect feature that is nearly completely unregulated and have attributed to numerous deaths and instances of psychosis.

As they have some niche areas where they excel will definately stick around longterm, but willl probably not be at nearly the capacity they are now once the bubble pops.

at least let us easily opt out of all AI features

I get what you say, and I do think that, but not the part where you say to delete Lumo, I have a solution for you: don’t use it, might be rude, but if you don’t like it or even if you hate it, just don’t use it.

Artificial intelligence and privacy are not necessarily in conflict.

Why do so many of the LLM haters want to remove it for everyone? Just don’t use it if you don’t want to. Just like I don’t use the wallet. I’m not whining about how it should be removed.

If you don't want to use AI, don't use/download Lumo. AI is the future, like it or not, and for Proton to not make it available with privacy in mind would be irresponsible and backward thinking. Proton is not pushing it in our face and it's not interconnected with all their other services.

Chat bot doesn't automatically mean privacy concerns though. However, allowing it access does allow for some pretty useful resources, especially if you're having trouble with something in Proton or if you need to find general information without using a resource like ChatGPT.

Instead, I think there should simply be a way to disable any AI usage for those that want it disabled.

Then they would have to read our emails to integrate the AI of course, to our emails.

I think this is important because of them slowly improving the chatbot "Lumo" they could pull a Gmail on us and say "Protonmail is entering the Lumo era"🤣🤣🤣