>>538054658
Different Anon here, but what metrics are you using to dispute the claim that a model specifically fine-tuned for storytelling is the best at its niche? That is where I feel you aren't reading anyone's words, or are deliberately choosing to nitpick them out of some peculiar sense of dismay that anyone could dare have a different opinion than yours. Especially when the dude already concurred with that instruct models are capable of outperforming it if wrangled properly? Not everyone wants to wrangle the smell of ozone away from their stories.
You perceive animosity where none exists. I, too, share his sentiment that NAI remains the best for its specific use-case.