Chat GPT 4 answers become very bland and short, kinda lazy

I read Devs are aware of Lazyness, the answer so far by Chat GPT are short, selfconclusive, few lines, void of valuable content and very bland and short .
Bug or feature?

1 Like

I would describe the responses of models, new since November, as being with responses lacking of substance. It’s like going TV shopping at Best Buy, and the 17-year-old clerk just meanders around the stuff he doesn’t know.

Short isn’t really a problem until you actually want something written that goes beyond a particular (low) token threshold. It is instead the chatty framing that is unneeded, repeating back to you for the sake of content that looked good for upvote button pressers and knowledge workers.

2 Likes