Philosophical Zombies: The Quest to Decipher Consciousness

The movie "Night of the Living Dead" now ranks among the classics in the horror film genre. Zombies, humans without apparent consciousness, loomed large in its plot, and in the terror the movie instilled.

But the concept of a zombie raises serious questions. Consciousness is decidedly a first person experience. If I know anything, it is my own conscious palette of sights, sounds, feelings, thoughts and desires. But though I can see others physically, and observe their outward behavior, I can not experience their inner life. So as sure as I know my own consciousness, I can not access that of others.

Are Other Humans Conscious?

We thus have a question of whether other humans are conscious. Are others just zombies?

On this question, I will cut right to answer. Others are almost certainly conscious. Though we can conceive worlds where "things" exist that look, act, talk, reason, eat, function, etc. just like humans, but don't have consciousness, we don't at all think that is the case on real Earth.

Why? Our fellow humans resemble us, and each other, too closely. All humans develop in the same way, we look and act the same, we share the same evolutionary ascendancy, we communicate with similar concepts, we grieve, we anger, we have brains organized in the same intricate manner, and on and on. Thus, even absent my ability to access your consciousness, our similarities are so great I would bet money that you are conscious.

Are Other Things Conscious?

Though I have no doubt you are conscious, the situation with a robot is not so clear. I can conceive, and in my lifetime could likely encounter, a robot-like computer that mimics human behavior so well its computer nature would be almost undetectable. It would converse with me, expresses emotion, declares preferences in music, and even, conceivably, has sufficient programming to paint (though maybe digitally, let's not get hung up on its mechanical capabilities).

Is the robot conscious?

We would think, in this version available in my lifetime - we will call it the first version - that the robot is not conscious.

Why do we judge that this first version robot lacks consciousness? We do so based on its internal architecture. For all their advances, the Intel chips (or any other mainstream chips) most likely at the core of our version one robot would not operate like humans. The chips would operate serially, mechanically, input-to-output, using speed rather than functional or architectural resemblance to mimic humans.

That is why we didn't for a second believe Deep Blue played chess the way Kasparov did. And that is why we don't believe computers of today's architecture have consciousness. Though they might mimic us externally, internally the architecture and configurations of current computers differ from humans like night and day.

What About in the Farther Future?

But let us consider a robot version two. We will dispense with traditional processing components. Gone will be registers, instruction stacks, memory caches, serial and parallel buses, data buffers, arithmetic logic units, and so on.

The second version robotic processing architecture will mimic the brain. We will have a neural network type architecture. Individual, simple units will connect in a brain-like mesh with thousands of other simple units, with long and short range connections, sub-communities of highly connected units, and dynamic weighting of connections as this second version learns. The entire collection will number billions of these units. In short, we will construct a brain in silicon.

Can we do this? Not now. Today, we have not reached a sufficient understanding of the brain and its networks, nor have we advanced non-traditional computer architectures enough, to accomplish this. But if we look a century from now, or maybe a millennium from now, we can conceive a world with a robot with a silicon brain.

Would it be conscious? I would not be surprised. As I see it, consciousness arises from attributes of the human brain and body. The mechanisms by which it arises may be complex and daunting. The properties of which it consists may be unique and unlike any other property of items in our universe. But whatever consciousness is, it is not random, it is not magic. We thus can conceive that humans can recreate the properties necessarily for consciousness.

Note I am not claiming the physical brain causes consciousness. The claim is that the brain has the conditions for consciousness. A concert hall does not cause a concert, nor does a flower cause a bee to fly to it. But build a concert hall, or place a flower, and the conditions are such a concert will likely take place in the hall, and a bee will likely fly to the flower.

Similarly, the human brain creates the conditions for consciousness. We, or at least I, thus have a reasonable expectation that we can figure our how to reproduce those conditions, and create a non-human entity that becomes conscious. This may require a century, or a millennium, or five millenniums, but mankind can conceivably achieve this.

Will We Know?

But will we know it is conscious? We will not, not for sure. The consciousness of our version two robot will be as first person as ours. While we think may have mimicked the conditions for consciousness, we might have just missed. Consciousness may depend on a dozen interwoven characteristics, and we may get only eleven correct in our robot. The robot may appear conscious, but be a zombie, and we couldn't tell either way.

To know, for sure, we somehow must connect consciousness to consciousness, to crack into its first person veil.

The movie "Avatar" might provide a clue. The movie features the transfer of the consciousness of our reluctant hero, Jake Sully, to his Na'vi counterpart body. It is interesting, and the movie puts enough techno-flashy ornaments on the transfer to make it believable, in a movie sense. But that is not a useful image here. Consciousness appears to be too highly dependent on a physical structure, putting aside religious beliefs, to be transported like helium gas or a winged spirit or an electrical burst from one physical entity to another.

Spock on the epic Star Trek series could connect mind-to-mind, thought-to-thought. That is a more useful image. In the Spock mind meld, consciousness doesn't float free of its substrate. Rather, it is through close physical contact that Spock breaks the veil and connects the consciousness of one sentient being to that of another sentient being.

But, alas, we have no idea what mechanism inside Spock's mind and fingers enables the connection.

Let's then go to version three of our robot, not a millennium from now, not five millennium from now, but a hundred millennium. Assuming, and this could be a false assumption, but assuming continued human development we enter a world where mankind has re-engineered itself. Humans are not born; they are nurtured to life as a hybrid biochemical, electrical, mechanical cellular beings.

And we now have multiple brains, each a center of consciousness. Even further, since the centers of consciousness exist inside a single "person", we have developed connections between these centers, connections that allow each center to observe and experience the consciousness of the other.

Is this conceivable? Without being ghoulish, we know now from Siamese twins that two centers of consciousness can exist in a single person. Siamese twins operate with two separate brains - and two separate centers of awareness, beliefs, emotions, intellects and so on - in the same body, and remain stable and alive for years.

Can we nurture an entity with multiple centers of consciousness, so they are linked? I will point to our rapid expansion of technological and scientific skills. Those skills leapt forward orders of magnitude in complexity in the last two centuries. A not unreasonable extrapolation would have mankind able to build sentient beings with multiple centers of linked consciousness within a hundred millennium.

Do we have hints of this? Yes. While dreaming, we can step outside of the dream, and be conscious of dreaming, while the dream continues, aka lucid dreaming. Blind sight, the phenomena where the brain perceives something without a top-of-the-mind awareness, also hints at multiple centers of consciousness.

Theses are faint hints, surely. I present them not to prove the case, but to provide some images. The case is that from these snippets our technological development can conceivably take us to a world where one consciousness can sense another; where we can pull back the first person veil of consciousness in a real life mind meld.

The first person veil of consciousness seems impervious now, but if I am correct that is not because consciousness is inherently first person. Rather, the veil exists due to our practical inability to sufficiently link to centers of consciousness. But our thought experiment of a far future world, where multiple brains exist inside one being, gives us insight. As we conceive two brains in one being, we can conceive of connecting them with the likely millions of neural links we might imagine required for one brain to experience what another brain is experiencing.

I have talked about millennium and hundreds of millennium before mankind can create the conditions for consciousness. Some circles of thought don't think it will take that long. Some predict conscious computers in decades. Hopefully we will have better resolved the enigma of consciousness before then, so we know the standing of such non-human consciousness. And hopefully this article, and the images it presents, adds a rung or two to the ladder of thinking needed for that resolution.

0 komentar:

Posting Komentar