By skyforbes 
This snippet of an interaction stuck out to me because it's the first time GPT basically asserted itself against me. I'm not sure if maybe I instructed it to do that or whatever in the past. Even in the past when stuff like this happens, GPT usually takes a "you're right, however…" or "yeah, I missed that aspect" kinda deal.
This time was "No, I heard you clearly the first time" and elaborated on how it already understood me the first time.
I've used GPT for a lot of things in the same scenarios and such and this is the first time I can recall this happening. Idk if anythings different or it's just a coincidence, but I enjoy this new model. I feel like it's closer to what I had to prompt the older models to be.