But nobody was arguing about the infeasibility of approximations of specific human cognitive abilities. Wasn't the conversation about general human cognitive ability?