I’ve decided that there is more to say about the irreducibility of consciousness.
How could the consciousness of something be tested? The Turing test was proposed this way; if I am having a discussion with a person and a computer, but I cannot tell which is the person and which is the computer based only on the discussion itself, then the computer must be a person.
I think there is a fundamental problem to this test; at most, it establishes epistemologically possible personhood, but certainly not ontological personhood. If I believe I meet a person, then at most I believe I meet a person, and short of tearing them open to see what they are composed of, I wouldn’t know whether or not they are a person or not (currently with our technology, its simply extremely unlikely that they wouldn’t be a person, but in the future, this could be a problem).
Ontologically speaking, what would need to be proven? My mere inability to know based on only one consideration (or even several) of the ontology is not enough for me to conclude about the total ontology; that’s quite a non sequitor. At most, the object’s ability to hold a discussion in such a way that I think its a person establishes that the object has the ability to hold a discussion such that I think it is a person. If it really is a person, then the explanation is granted.
To see why I would hold that the ability to discuss does not lead to personhood, see this thought experiment.
I design a “Chatbot.” This Chatbot is meant to hold discussions with people by figuring out their topic of discussion (say, by picking out some keywords) , do a google search for that topic, and then summarize a part of the first link on the cogent issue. It would also have some relevant programming on conversational courtesies.
I decide to demonstrate my Chatbot by holding a conversation with the Chatbot in front of a crowd. I say “Hi.” The Chatbot recognizes my “Hi” as a salutary greeting, and based off of stock replies programmed into it, replies “Hello.” I offer a hand, and say “My name is Bryce, what is yours?” The Chatbot offers out its hand (because of the program) and says “My name is Chatbot 1.0. Nice to meet you Bryce” (again, because of the program).
This entire sequence here on the part of the Chatbot has been determined by its programming; its done nothing outside of its programming through its own volition. All that it did is essentially comparable to the mechanical “programming” of a mechanical calculator, only of a more complex manner.
Think back to the Turing test. How can it be concluded that the computer must a person, when it could also be concluded that it is no more than my thought-experiment Chatbot?
How then could the consciousness of an object be proved, then?
First, consider humans. What proves our consciousness? I would say that it is our feeling of consciousness; that I would feel free will but not really have free will seems an absurd idea to me. Further, I feel qualia, and qualia is irreducible to to matter as well.
So what proves our consciousness? An irreducible component to our thinking and feeling. There are some parts of us that are undeniably reducible; for instance, I feel hunger because evolution found that it is better for a creature to feel hungry when it hasn’t eaten for such-and-such time, etc. However, there are parts of me that are irreducible as well; the feeling of hunger in the first place, or my feeling of free will. Any attempts to trace these back to material processes ultimately backfire and lead to absurdity (either by leading back to a denial that we have qualia in the first place or that somehow the determinative process can include creatures who feel non-determined).
Computer consciousness could never be proven because there is no computer that exists that isn’t programmed. Any computer that displays an aptitude for discussion or the ability of analyzing colors is really not doing so out of any un-programmed ability, but leads back to the program which tells it what outputs to give to certain inputs, essentially comparable to the input-output of any other mechanical device.
The Turing test then fails because if I were allowed to take into consideration other data, then I would be able to see that the computer’s ability to discuss is based on a complex program, not because of any irreducible component.
