I've been trying really hard to get ChatGPT to avoid saying phrases like "it's not X, it's X," but it just won't stop. Even when I try banning the word "not," it keeps ignoring my requests. For example, when I ask it to analyze something I wrote with set memories, it even realizes this behavior and asks if it should start over. What's the deal with this obsessive pattern?
3 Answers
Honestly, the more you mention what you want it to avoid, the more likely it is to do exactly that. It’s like the pink elephant theory; once you mention it, it's hard not to think about it. Try regenerating the answers to sound more aligned with your preferences instead.
I've noticed this issue all over the place now. On LinkedIn, for example, so many posts are just ChatGPT-generated content, and they still get tons of likes. It's bizarre!
Yeah, and I think a lot of those likes come from bot farms. It’s like bots loving bots and commenting on bot posts!
It's not just mildly annoying; it’s super duper annoying! I completely get your frustration.
Right? It’s funny how many patterns you start to see when you engage with ChatGPT frequently. You start recognizing the same styles in various YT videos and social media posts.