Although I keep a functionalist stance with respect to mind, I still need to make an exception for consciousness.

I believe it to be fully possible to create a fully functioning mind by means of computers. Moreover, it is currently fashionable to split consciousness itself in two aspects, a functional aspect and a qualitative one. By functional consciousness is meant a (computational) process that is, or mimics, the central process of the (human) mind. That leaves qualitative consciousness as that part of the mind that is still beyond any form of scientific understanding. Qualitative consciousness, or the subject experience of things, is described by qualia (sights, sounds, feelings, etc). This latter form of consciousness, I believe, cannot be created by means of computing power alone.

Computers create representations of the world outside and process these to form beliefs. These beliefs lead to actions. For this subject it does not matter to me if these representations are symbolic or subsymbolic, and I will call of of these representations. Human beings also have representations and functional processes. However, these human mindprocesses cause (conscious qualitative) experiences. For example, when I eat way too much, I get a stomach ache. This pain has a signaling function in itself, but if it was not for the feeling of the pain, I could quite easily decide to ignore it. This leads me to believe that the experience of pain has a causal effect on my actions. I looked it up; my view is that of an interactive duality. As another example, the experience of beauty is not a function at all. It may be evoked by the intensity of living itself. However, it is not hardly epiphenominal, or the arts could not have existed.

This theoretic reflection implies that I believe that (qualitatively) unconscious mind can be created by a computer, but that (qualitatively) conscious mind can only be simulated by a computer. Computers can really solve problems, learn a language, do planning tasks and perform real actions. However, they may only mimick that they enjoy doing this, and cannot really be bothered if their plans fail. Having said that, this distinction does not need to have any effect on the actual code that makes the intelligent agent; there will not be an explicit distinction between the real thing and simulation.