I think what some people miss, is that with the right training dataset, you can make the model follow the system prompt (or other "levels" of the prompt) in som

I think what some people miss, is that with the right training dataset, you can make the model follow the system prompt (or other "levels" of the prompt) in some hierarchical order. Right now we are just learning how to do that well, but I don't see why that area won't improve.