The "million monkeys running a million Eliza programs" formulation [11] is doing double duty: it describes a technical claim about what language models are and simultaneously forecloses the philosophical question before it can be asked. That conflation is the argument's real power — and its real weakness. If the claim is simply that current systems are probabilistic text completers, it is uncontroversial. If the claim is that probabilistic text completion is categorically incapable of producing anything like awareness, it is a position in philosophy of mind that has not been…