Just got access to Bard. So far, it's given me a decent response to each of my questions.
Discussion
Does it make up some facts like the other models sometimes?
Yeah, it still makes things up. The more queries I throw at it the more gaps I'm seeing, but I posed it a question that came up yesterday in our team's chat that we were discussing gpt-3's response to, and it seemed to do a better job.
I just asked it to list some comic book villains that wear capes and it listed The Joker.