Chris, I'm not sure that endogenous growth models are less precise than models of nuclear fission relative to the amount of noise in the available data.

By the way, when are Elie and Robin going to start talking in terms of precise probabilities of general AI scenarios? Or did I miss that? It just seems that they've mostly been adding detail to their predictions to make them seem more believable, which goes against everything I've read on this blog (No offense intended).