> We have people here who use LLMs to raise their children, to manage relationships, to design products.

My first thoughts were around trying to understand why these people would do that. What I see a lot around me, is people being afraid to fail, as it's some inherently dumb and bad thing, not realizing that sometimes failing is what makes you learn enough so you don't fail later, or builds resilience in other ways that later will be useful, for others or for yourself. Avoiding the path of harsh failure will put you down the path of mediocrity.

It's passing the buck, isn't it, if I ask AI and AI tells me something dumb, I can blame AI and absolve myself of responsibility.

At the same time, the troubling allure is that the machine has ingested a million books and has better knowledge than me (e.g. in things like JavaScript or Japanese), so why not trust it about how to raise kids.