Comment by cootsnuck

Well, it seems like collectively we are all struggling to perceive model progress, given that it seems like every reply to you is reporting different experiences with which of the models has subjectively performed best for them.