Bayesian reasoning isn't a fallacy. A known expert in one domain is often correct about things in a related one. The post didn't claim that Carmack is right, just that that he's who they would bet on to be right, which seems perfectly reasonable to me.