From my conversations with ChatGPT, it seems like it's still a long way off before AI replaces humans. I'm not saying it's useless, but "for all its sophistication," it still plays the role of a novice assistant.
Comments
Log in with your Bluesky account to leave a comment
When I ask ChatGPT for an idea, it gives me it and I critique it. When I give it an idea I've come up with, it rarely critiques it, instead offering near-absolute agreement. It's certainly not afraid, but it still doesn't know how to critique.
For example, I asked it about scenarios for the future of Egypt after Sisi, and it gave me 4 scenarios. I critiqued them and excluded one of them for being unrealistic. When I presented 4 other scenarios of my own, it agreed with all of them without any criticism 😅
Comments