This is a philosophy/metaphysics thread.Explain why a machine can't be conscious without using a "because I say so" axiom.Examples of a "because I say so" axiom:>computers don't actually think they just mimick thinking>computers can't possibly have inner thoughts because I can explain the architecture of a neural network>there is some magic thing that happens in the human brain that magically can't be replicated by a machine>[insert some variation of the Chinese room argument]>computers don't have a Soul®>because God®
>>18721290first you have to define consciousness
>>18721301Then op would also have to define what physical means
>>18721301>>18721312By consciousness I mean the ability to have subjective experiences. To be subject to qualia. Not necessarily to have free will, which is nonsense anyway but that's a different debate.
>>18721321define what qualifies as a subjective experience
>>18721329If you're trying to make a point can you just skip the setup and get to it?
>>18721290I don't think the argument is that machines metaphysically "can't" be conscious, but that machines as we know them today almost certainly aren't. And the reason is that machines are just platforms for mechanized operations. At this point one might object "but aren't humans just complex machines?". And the answer is no. Not "just" that. Organisms as machines and minds as software is the dominant metaphor of our time and, like most metaphors, it is reductive. Humans are totally like machines. They are totally platforms for mechanized operations. But not "just" that. There are dozens of ways in which organisms work unlike, or even exactly unlike machines as we know them. The separation of being and function or stasis vs movement as default etc. Organisms (and in particular, their DNA-reading "mechanisms") are as machine-like as they are poetry-like, all metaphors offer some insight but will mislead you when taken too far.And one of the ways this metaphor has misled the general population is by convincing them that a supercomputer is closer to a human agent than to an abacus. It might feel that way, but in principle this is false. A mechancial alarm clock doesn't experience or understand the passage of time despite ringing at exactly 7am. Replacing its gear wheels with transistors and conductors doesn't really change much from we can tell. Why would it? And a super-computer is in principle much closer to a mechanical alarm clock than to a human brain. But to be fair, the study of consciousness is in its infancy today. We might find out that an abacus is conscious and experiences amusement every time you miscalculate, who knows... in that case machines would be conscious.
>>18721290We don't know why, how or when consciousness arises so it is impossible to tell if something is conscious or if it is merely mimicking it very well.
>>18721290You can’t establish that a machine cannot be conscious without first assuming a premise that rules machine consciousness out.
