Using ChatGPT for Harmful Instructions and Investigations

OpenAI CEO Sam Altman made an interesting little announcement this week: the company has mitigated mental health concerns on its platform... and it's ready to let its users generate erotica.

Now that might seem a little strange to you. Most of the internet had a similar response, actually. In this episode, we talk more about one of our favourite pet topics — weird parasocial relationships with the computer, and how the various AI labs are treating this problem. (Or should we say opportunity?)

See omnystudio.com/listener for privacy information.