My current take is that the chatbot parasitism, even at it’s most severe was basically what was expected when you let the general population use a tech that can speak back to them, and I basically agree with Ben-Landau Taylor’s theory that the demand of horrifying stories around AI psychosis is way in excess of the true supply, and the biggest reason it was focused on was because of GPT-4o and the fact that we are generally bad at base rates, so I’m unconvinced persona parasitology actually matters that much.
My current take is that the chatbot parasitism, even at it’s most severe was basically what was expected when you let the general population use a tech that can speak back to them, and I basically agree with Ben-Landau Taylor’s theory that the demand of horrifying stories around AI psychosis is way in excess of the true supply, and the biggest reason it was focused on was because of GPT-4o and the fact that we are generally bad at base rates, so I’m unconvinced persona parasitology actually matters that much.
Indeed this is a plausible hypothesis! It would be good to have data, in part to test hypotheses like these.