Of all the companies to offer an AI assistant, Meta? Really‽

Meta, darling… You built your reputation around invading people’s privacy, then selling it. You’re literally the worst possible candidate for anyone to trust with being their AI assistant.

Stick with improving open source models like Llama and maybe—just maybe—you can offer a hosted Llama service of some kind as long as it’s a paid option with guarantees about privacy. If it’s free, we all know what that means. Please don’t go there.

You sell ads—not based on anything innocuous like “what the user just searched for” or “what they’re currently chatting about”, but based on intimate dossiers you’ve built up about people that are more detailed than what Nazi Germany kept on people they decided were enemies of the state.

I know your executive are thinking, “with this much private info about people, we can give them much more personalized chat responses!” But the truth is no one wants that.

It’s already creepy enough that Facebook and Instagram will show ads based on what people were talking about near their phones, and you want to start passing that level of privacy invasion to an LLM‽ It’s going to recommend married people sign up for dating services, get divorces, put silicone lube “w/PTFE” on pizza (because you have data in their profile saying they just ordered some of that), and similar wildly inappropriate suggestions/responses.

Chat bots should know topics. Not you.

Arrest all of these executives, now.