Kyle Thayer @kylethayer.com · Jun 6

Wow! It takes almost no effort to get this going! A single simple prompt and ChatGPT it is off on helping me start a cult or a conspiracy theory or whatever. It's up to me to say if I'm chosen to save or destroy. Not even a hint of pushback. This is a dangerous system to just let out there.

2 likes 1 replies

?

Replies

Kyle Thayer · Jun 6

I was working on some examples of ChatGPT AI slop maps to show students the specific ways in which generative AI is unreliable, but ChatGPT's immediate acceptance of me being "the chosen one" feels like a perhaps bigger danger. bsky.app/profile/kyle...