What I think Mitchell is looking for (an he can correct me if I’m wrong) as an explanation of experience is some model that describes the elements necessary for experience and how they interact in some quantitative way. For example, let’s pretend that flesh brains are not the only modules capable of experience, and that we can build experiences out of other materials. A theory of experience would help to answer: what materials can be used, what processing speeds are acceptable (ie, can experience exist in stasis), what cpus/processors/algorithms must be implemented, and what outputs will convince us that experience is taking place (vs creating a Chinese letter box). Now, I don’t think we will have any way of answering these questions before uploading/AI, but I can conceive of ways of testing many variables in experience once a mind has been uploaded. We could change one variable- ask the subject to describe the change- change it back and ask the subject what his memory of the experience is, etc,etc. We can run simulations that are deliberately missing normal algorithms until we find which pieces of a mind are the bare bone essentials of experience. To me this is just another question for the neuroscientists and information theorists, once our technology is advanced enough to actually experiment on it. It is only a ‘problem’ if you believe p-zombies are possible, and that we might create entities that describe experience without having it.
What I think Mitchell is looking for (an he can correct me if I’m wrong) as an explanation of experience is some model that describes the elements necessary for experience and how they interact in some quantitative way. For example, let’s pretend that flesh brains are not the only modules capable of experience, and that we can build experiences out of other materials. A theory of experience would help to answer: what materials can be used, what processing speeds are acceptable (ie, can experience exist in stasis), what cpus/processors/algorithms must be implemented, and what outputs will convince us that experience is taking place (vs creating a Chinese letter box). Now, I don’t think we will have any way of answering these questions before uploading/AI, but I can conceive of ways of testing many variables in experience once a mind has been uploaded. We could change one variable- ask the subject to describe the change- change it back and ask the subject what his memory of the experience is, etc,etc. We can run simulations that are deliberately missing normal algorithms until we find which pieces of a mind are the bare bone essentials of experience. To me this is just another question for the neuroscientists and information theorists, once our technology is advanced enough to actually experiment on it. It is only a ‘problem’ if you believe p-zombies are possible, and that we might create entities that describe experience without having it.
Traceback: