If living creatures, such as ourselves, are evolved biochemical mechanisms, and should free will exist, what natural neurophysiologic phenomenon could possibly give rise to it (that would not be as deterministic as, say, any other chemical process)? And if we are indeed biochemical structures (as biologists in general believe), why might not appropriately designed future machines (advanced AI) likewise have the capacity to exercise free will (should free will exist)?

Don't forget the compatibilist account of free will (see the entry here), which says that we can make free choices -- i.e., choices for which we're responsible (including morally responsible) and properly subject to praise or blame -- even if our choices result from totally deterministic processes. In other words, free will doesn't require the falsity of determinism. I know of no cogent arguments against the compatibilist account of free will.

According to compatibilism, then, we can make free choices without needing any mysterious, non-causal, or indeterministic neurological goings-on. By the same token, an advanced AI machine could also make a free choice, provided it's advanced enough to be able to entertain, appreciate, and evaluate reasons for and against making (in its own right) some particular choice and able to choose in accordance with that evaluation. As far as I know, such machines are a long way off, but I see nothing in the concept of free choice that rules out, in principle, their making free choices.

Read another response by Stephen Maitzen
Read another response about Freedom, Biology