the perfect model as I often think of it like what is at the end of the RSI rainbow. I I think I think it's very likely that we will discover what the perfect AI model looks like. The most perfect use of flops or ops or whatever the operative unit of compute ends up being. It will be globally mathematically provably optimal. And at that point, I think we'll know whether we need to keep doing this sort of horizontal exponential buildout and build the Dyson swarms or whether it's just so perfect, so mathematically
I I think I think it's very likely that we will discover what the perfect AI model looks like.