Okay, hear me out. I love GPT-4, but lately, I've been playing with Claude 3... and it's blowing my mind with how accurate and helpful it is.
Is anyone else noticing this? Claude 3 seems to nail instructions way better than GPT-4, and its answers feel a lot less prone to... creative tangents.
Makes me wonder:
Is raw creativity less important than we thought? Could GPT-4's focus on that actually be hindering its usefulness in some cases? Curious to hear everyone's thoughts!
[link] [comments]