If it's a mistake, it should course-correct.

I agree that some of the smarter models are actually worse. I hope they take a model that's good enough--there are many--and just try to get it chatjimmy.ai speed.

I have to think that's the future, somehow, and I'm really excited about it.