Can there be a human who goes to work, has a wife, coaches little league, but has no conscious experience? Perhaps here we have a way to investigate the possibility. Does Ambien shut off excitatory (or activate inhibitory) neural circuits that directly disengage a "consciousness" module or diffuse neural net? Can a somnambulant Ambien user "learn" new things, be they cognitive or motor skills, and can these new skills be performed/recalled at a high skill level when fully conscious, despite a lack of mnemonic recall for the event itself? Or can he/she be made to become fully conscious while in this state e.g. through the use of pain or loud noise, or direct stimulation of other neural circuits? To what extent does Ambien actually interfere with conscious perception? Do Ambien users display their normal personality when they are having a somnambulatory event? Can we stick them in an MRI or EEG and observe how activation patterns differ when the person is conscious and when they are on Ambien?This is a brilliant connection. People who take Ambien are being found to walk, eat, even drive without apparent awareness or memory, and this certainly looks a lot like the philosophical zombie. Zombified Ambien users are clearly reacting to the outside world -- for instance they avoid obstacles, at least to a certain extent, when they drive -- but they don't seem to be aware of it. In other words, when they see something they don't know that they see it, and they don't form memories of seeing it (or of knowing that they saw it). Looking at the differences between someone's neural reaction to a stimulus on Ambien and their unimpaired reaction could give a lot of insight into what it means to be conscious of something (and by extension, what it means to be conscious, period).
Neurotopia speculates that "David Chalmers must be wetting his pants," but I don't know. As I understand it, Chalmers wants the possibility of zombies to mean that consciousness can't be tracked to an observable physical process. Presumably, if it's possible for a zombie to exist, that's because it's possible for there to be a creature that is physically (including neurally) identical to a human but that lacks conscious experience anyway. This essentially makes any study of Ambien users inadmissible for Chalmers' purposes, though of course he'd be able to explain the findings away -- see my earlier comments on refuting pseudoscience with facts. If there turns out to be a measurable distinction between a "zombie" brain and a fully conscious one, that shows that there's some neural basis for consciousness, but the beauty of Chalmers' take on the "hard problem" is that this isn't enough to prove consciousness to be purely physically based. Nothing is.
So these guys are good news for neuroscience, and we should definitely chain them up in the shed behind the MEG lab. But there's no refuting a philosopher so spooky-minded that he wants to believe in zombies. Seriously, zombies. I ask you.