Id like to propose Altman’s law: the amount of energy needed to refute bullshit generated by linguistic AI is two orders of magnitude higher than it takes to create it.