Books on Amazon
McGinnVsTuringtest: so leveling it may seem, it is a poor criterion for machine's own consciousness:
1. It provides no necessary condition for consciousness, because many conscious beings would fail miserably in the Turing-test. (E.g. cats, which cannot speak but have a consciousness). The test is too strong linguistically oriented.
2. The criterion is moreover not sufficient: the test is obliged to a behaviorist conception of consciousness.
Behavior/consciousness/McGinnVsBehaviorism: consciousness states cause behavior, but they are not reducible to behavior. Evidence, based on behavior, is always fallible. E.g. an actor can pretend to be in pain. In this way, behavior testifies of consciousness, but it does not guarantee it.
McGinnVsTuring-test: tries to tell when something has consciousness, without saying what consciousness is. It could be that we could build a machine that passes the test, but the darned thing would not have the slightest consciousness.
Therefore calculators are possible, they do our work without consciousness.
We could even assume that any cognitive task that we do with the help of intelligence, understanding and consciousness, could be done by a machine without intelligence, understanding and consciousness. ((s) Thus the poor behaviorism triumphs).
Each piece of rational mental activity could thus have a computational counterpart. (> Danto).
The counterpart reaches through a not mental way, what we achieve through a mental way.
If we can do in principle X without having Y, then, the fact that we can do X proofs not that we have Y.
Die Grenzen vernünftigen Fragens Stuttgart 1996
Wie kommt der Geist in die Materie? München 2001