What Things are Conscious? | Eric Schwitzgebel

Eric, consciousness is a passion to understand that I've had, that you've had, and we approach it philosophically, neurobiologically. Another approach says stand off and ask the question, what are the kinds of things that are or could be conscious? We think, we know, we're conscious. Maybe higher animals are. What about computers? A lot of argument about computers, or the internet can be conscious. What is a methodological way that we can array the landscape of possible things that could be conscious and an approach to analyzing it? So I'm skeptical that we can know what things are conscious, right? So I know I'm conscious. I'm not sure about you though. I'm highly confident that you're conscious, right? You're a lot like me in so many different ways, right? But once we get beyond that, I feel like I start to lose my grip. So I think the range of possibilities is immense. So on one end, you have panpsychism, which holds that everything is conscious, all the way down to individual particles. And I think that there's some beauty in panpsychism. This has been being defended recently by Galen Strassen and David Chalmers and others. On the other end, you have a view on which maybe only human beings are conscious, or maybe only human beings in their more reflective moments, right? Maybe in those self-reflective moments when you're thinking, ah, here's the visual experience I'm having, then you're conscious, right? But if you just go through your day not reflecting and just spontaneously responding, maybe you're not, maybe you're not conscious during some of that period, right? So there's the distracted driver case, right? You're driving down the freeway, you're thinking about such and such, right? You get there, you don't even remember that you were even driving. So you might not have been conscious of the road at that time, right? Maybe you had a stream of thoughts, or maybe not, right? It's possible that during some of those moments, it's entirely blank. I don't know that I could rule that out entirely. So these are two pretty extreme positions. I'm not particularly attracted to either of them, but just as bookmarks, I think they're both still open as possibilities, right? And if those are open as possibilities, then saying, yes, here's how we figure out whether frogs are conscious, seems, or here's how we figure out how computers, or under what conditions computers would or would not be conscious, that's, I think if you're open to these other two possibilities, those kinds of questions start to seem like, wow, how could you answer those specific questions? I don't think we have a good metaphysical theory of consciousness, I don't think we have a good scientific theory of consciousness, until, and without them we can't address questions like whether a computer is conscious, or whether a frog is conscious, and in order to develop a good scientific or philosophical theory of consciousness, it seems like we need to make assumptions about whether a computer would be conscious, if designed in the right way, or whether a frog is conscious. So I'm inclined to think that you've got kind of a tight little circle here, right? In order to evaluate one of those questions, you need a theory of consciousness. In order to develop a theory of consciousness, you have to make presuppositions about how to answer those questions. Now maybe we can break out of this circle eventually, but I don't think we're very close right now to breaking out of that circle. So I'm pretty pessimistic about our ability to come up with really good answers to these questions now. Take a specific entity, a computer is one that is popular today, that among many people it seems intuitively obvious that eventually we will create a computer that will be conscious. There's no doubt we will create a computer that will act as if it's conscious, and could convince anybody who tests it that it is conscious, and will complain if you say that it's not conscious. But the fundamental deep question is, is there an inner sense that that computer will have? So I am inclined to agree with what you just said, that if technology, we don't blow ourselves up, and if technology continues to progress, that eventually we'll create some sort of artificial entities that exhibit outward behavior that seems very natural to interpret as indicative of consciousness. Now my inclinations are skeptical, so I'm not entirely sure whether we should say such a thing would be conscious, but my inclination is to think that the safer default supposition would be to attribute consciousness in those cases. I guess there are two reasons to think this. One is, I suspect that would be the natural reaction of people, the kind of folk reaction, and that carries some weight with me, although I also think that folk reactions and folk opinion can be very wrong, so I don't think it's decisive. But that has substantial weight. And the other thing is, I think if we're going to make an error about such beings, it's better to err on the side of attributing them consciousness. I think it would be a moral horror to say, well, we don't know whether machines are conscious, we're going to treat them as non-conscious, and so we can enslave them. And then it turns out, really, metaphysically, or at the root, they really are conscious, that would be horrible. So it seems like if you're going to make a mistake, the mistake to make is to over-attribute rather than to under-attribute. From a moral point of view. Not a metaphysical point of view, but a moral point of view. Right, so metaphysically, I'm somewhat skeptical, I have my inclinations that tend to probably favor that to some extent. Well, that's a materialistic view. I think the capacity to generalize consciousness is based on a materialistic view of consciousness, which may, in fact, be the case. But I think it would be inconsistent to say you have a materialistic view of consciousness and that to have consciousness in a computer is impossible. I don't think that's a consistent position. Right, so it might partly depend on what a computer is, right? So you might think, well, it's impossible to have consciousness in silicon chips arranged in a certain way. But if by computer, you just mean some sort of artificial human construction, so that you could even have maybe an artificial biological construction, then it seems very natural to think that materialism is going to lead you to attributing consciousness to these artificial systems. And if you have a definition of consciousness, which some people do, which have to do numbers of interactions per second or complexity of information or something like that, I mean, there are arguments that some have propounded that says, you know, it's possible, conceivable that a neutron star is conscious, because there are so many different calculations that can be done per second in some very sophisticated way. So I think there's something very elegant and attractive in the idea that wherever you have information or sophisticated or complex information processing, you've got consciousness. But also, it seems a little strange to say that a neutron star is conscious. And you might think that what you really need is a certain kind of goal-directedness, which doesn't seem like neutron stars have, or a certain kind of way that it's structured or hung together that a neutron star doesn't have. And maybe that's true. So I think that it's, I'm inclined to think that it's hasty to say consciousness, information, neutron stars. But I also think that it's hasty to dismiss that, because probably something is true about consciousness that seems strange. When you look at the history of views about the metaphysics of consciousness, every single view in the history of the philosophy of mind is bizarre. It's got some bizarre implications. I mean, you look at, Descartes was bizarre. He thought dogs didn't have conscious experience, right? Some people think dualism is intuitive. But any person who's worked out a detailed dualist view ends up with bizarre implications. Materialism has bizarre implications, like maybe that on one version, maybe neutron stars are conscious. On another version, maybe you've got group consciousness cases. On another version, maybe you've got consciousness in weird kinds of computer systems. All the views have strange implications. So when you're in consciousness, no solution to the problem will give you something that has a simple elegance to it. Everyone will have consequences which are bizarre to our normal understanding. That says a lot about consciousness. I think it does. That there is no solution to consciousness that does not have bizarre implications. I think that's, the empirical evidence suggests that, right? Attempt after attempt for hundreds of years to address the mind-body problem has resulted in every single well-developed theory having some bizarre implications, right? If it were possible to construct a non-bizarre theory of consciousness, I think probably someone would have already done it, right? So the answer to the question about how consciousness arises in the universe or the answer to the mind-body problem, I think is inevitably going to be bizarre. My inclination is to think that that's because our folk intuitions about consciousness form an inconsistent, incoherent set, right? And so it's inevitable that any well-developed view is going to violate them somewhere. But the problem is that we have to rely on our intuitions to do philosophy. So we're relying on an inconsistent set of things to try to answer the question. And so we end up in a tangle.