KI-Ethiker warnen vor „digitaler Spukerei“ durch tote Angehörige

I saw this posted elsewhere and thought I’d bring it up to talk about the idea of AI chatbots that for a lack of a better word, „deep fake“ loved ones?

Too creepy? Or a neat way to plumb your parents previous utterances and writings to create conversations, etc.?

Coming from somebody who lost my parents as a young adult, it would be interesting to have „discussions“ with a LLM trained with my parents correspondence, etc. I assume it would create the illusion of their style and knowledge. Might be comforting, but also might be unsettling. Of course, they passed 30+ years ago, so there’s not a ton of training material I could find at this point.