Comment by zzzeek

OpenAI's training is basically the entire internet. And we know when it comes to deciding if a story really is "that bad" or if it's hyperbole, the internet is really good at giving us a clear picture on that. Police shootings? ah well he was acting funny, they had to shoot him. School shootings? so many more people die of cancer each year. Fascist / authoritarian leaders on the rise? Here's 5 subtle and pointless ways Hitler was different and therefore NOTHING TO WORRY ABOUT. Global warming? The jury is still out. "Reasonable" discourse on campus being "cancelled"? OH YES CANCEL CULTURE IS RAMPANT!

Yeah, AI can totally decide what's real and what's not. Sounds great