I am not a patternActually there won’t be mortality by the end of the twenty-first century. Not in the sense that we have known it. Not if you take advantage of the twenty first century’s brain-porting technology. Up until now, our mortality was tied to the longevity of our hardware. …As we cross the divide to instantiate ourselves into our computational technology, our identity will be based on our evolving mind file. We will be software, not hardware.
– Raymond Kurzweil, The Age of Spiritual Machines, p. 128-129 I would dearly like to believe that once I’ve abused this body to death, I’m going to be able to change shells like a hermit crab. But it’s not going to work. I’m going to die and so is Ray Kurzweil. Our bodies are going to rot and we won’t be around to watch them from within our new silicon homes. The reason is simple: Consciousness is not a pattern. It’s not even a wildly complex pattern. The real question is why we would ever think that it is.
Let’s imagine that we accomplish all that Kurzweil says we have to in order to make possible the porting of our consciousness into a computer. We are able to map all 100 billion neurons in the typical human brain. We’re able to map their 100 trillion connections. We learn the rules by which the brain operates: stimulate this nerve in my finger and we can predict which neurons in my brain will light up. We build computers big, fast and cheap enough to accommodate all that data and to run the “brain program” that instantiates the rules we’ve learned. All of this seems possible within the century and perhaps even by 2029 as Kurzweil predicts.
In short, we will have created a perfect model of the brain. It undoubtedly will be highly useful since it will allow us to perform experiments that were previously impossible for practical and ethical reasons. But the computer running the brain program on that data is not itself conscious.
How can we be so sure? After all, as Kurzweil writes:
Is he – the newly installed mind – conscious? He certainly will claim to be. And being a lot more capable than his old neural self, he’ll be persuasive and effective in his position. We’ll believe him. He’ll get mad if we don’t. (p. 126)At the very least, shouldn’t our presumption be that the computer that’s belligerently insisting that it’s conscious is indeed conscious? It’s certainly passing the Turing Test.
But there’s a reason why we shouldn’t be fooled by this silicon squawk box, no matter how silver-tongued its line of palaver is.
First, we need to recognize that we can be fooled by a non-sentient being into thinking that it’s sentient. This actually happens on occasion. For example, some people were apparently fooled by Joseph Weizenbaum’s Eliza program that parodied Rogerian psychoanalysis. I have been fooled by the first few responses of automated responders on telephones. Kurzweil’s own company wrote a program that composed and performed original works for a jazz quartet that I thoroughly enjoyed even after I found out that they were computer-generated. It is completely possible, therefore, to think that something is conscious even though it isn’t.
But, in each of these cases, a little further investigation could expose the artifacts for what they were. Perhaps a few more questions would expose Eliza for the simple parser and small array of responses that “she” is. True, but that’s not the point. What I’m saying is more obvious than that: Just because something seems to be conscious doesn’t mean that it is. A mechanism that fools us into thinking it’s conscious for a longer time isn’t more conscious than one that fools us for less time. It may just mean that we’re bigger fools.
If consciousness is not the same thing as the appearance of consciousness, then what’s the secret ingredient that’s missing? That’s tough to characterize. John Searle in his famous Chinese Room thought experiment says that what’s missing is intentionality, but that is a word with a long philosophical history, and Searle’s interpretation of it is difficult and contentious. So let’s not even try to characterize the “thing” that you and I have that Eliza does not. Instead, let’s look at one thing that that ingredient isn’t.
Imagine the computer that has replicated Kurzweil’s brain state and is running the brain program. Call it KurzByte. We’re talking to it and it’s responding. The voice module they’ve hooked up even sounds just like Kurzweil. What makes it conscious, though, isn’t the sounds that come from the speaker; no one thinks that a recording of Kurzweil is conscious. What makes KurzByte conscious is what causes those sounds to be emitted. If we were to change the sounds – turn down the volume, adjust the pitch, slow the clock speed – KurzByte would still be as conscious as ever. What makes KurzByte conscious is the pattern of neural activity it’s replicating. KurzByte got to be the new, immortal Kurzweil by having a one-to-one match-up between neurons and data.
But that correspondence is only there because we choose to take there to be a correspondence between the neurons and data. So, if KurzByte is conscious because it replicates neural patterns , it is conscious only because we take it as so.
After all, the universe is an awfully big place. I’m quite confident that somewhere in some galaxy there is a gas cloud with 100 billion molecules in the same pattern of ons and offs as the bits in KurzByte for some stretch of time. In fact, we get to define what constitutes an on or off in a gas cloud: perhaps clockwise rotation is on, or perhaps moving above a certain velocity is on. That gas cloud is not conscious even though it is identical to the shifting set of patterns in Kurzweil’s brain during some stretch of time. How do we know this? Because we could decide to take the pattern of molecules differently. Perhaps we’ll take clockwise rotation as off. We’re perfectly free to do that. In such a case, whether the gas cloud is conscious or not depends on how we take the pattern. You might take it one way and I may take it another, in which case the gas cloud is conscious and not conscious at the same time.
But that violates what seems to be the most fundamental fact of consciousness: I’m conscious whether anyone else knows it or cares. Kurzweil – and all of us – aren’t conscious simply because someone is taking our neural state one way instead of another. Consciousness isn’t simply a pattern, although certainly there are patterns we can observe in the neural states that give rise to consciousness.
So, what is consciousness beyond the patterns we can observe in our neural states? I don’t know. I’m not smart enough. But let me make a gesture in a direction. You will, I hope, find it trivial and obvious.
Take a game of pool. It can obviously be described quite precisely in terms of locations and movements. Put that data into a computer. Or just take micro-measurements of the starting positions, the table’s properties, and the force imparted to the cue ball at the beginning of each turn. Add a program that models Newtonian physics and you can compute an exact replica of the game. Now play the game by printing out the turns in text, by displaying a high-resolution visualization of it, or by programming the computer to vocalize a play-by-play narration in sportscaster style. Do you count that as playing pool? Does the pool game now live forever in the computer?
I can’t even make sense of that notion. What we’ve put into the computer is a representation of a pool game, not a pool game. Likewise, a choreographic score is not a dance. A description of the exact placement of daubs of color is not a painting. And a digital representation of something isn’t the thing itself.
But if the difference between KurzByte and Kurzweil is the type of matter the Ur-KurzPattern is pressed into, what makes Kurzweil’s matter so special? Beats me. We call it “flesh” and it seems to be a miraculous sort of stuff. It doesn’t just represent a pattern of life. It’s not alive only if we take it that way. It’s actually alive.
What does that mean? What is life? Again, I don’t know, but I don’t have to know to maintain that a representation of it isn’t itself alive.
But, does it matter? Why not let Kurzweil model his brain in silicon if he wants to? And of course we should. Going down this path will teach us much about how the fleshy brain works. But even when KurzByte is pleading with us not to reboot “him” even though it’s to install a faster processor so that he can live Kurzweil’s life 50% faster, we should remember that KurzByte only sounds conscious.
More important, for two reasons we should maintain our skepticism about the extreme idealism KurzByte’s proponents would have us accept as obvious.
First, the KurzByte sort of extremism leads to conclusions we don’t want to accept. It declares two things (Kurzweil’s brain and KurzByte) to have the same property (consciousness) if they have the same formal property (pattern), but we are free to find patterns wherever we want to – gas clouds, counting every fifth grain of sand on the beach, clustering leaves on a tree by which way they’re blowing or by the intensity of their coloration. Thus, we can find consciousness in places we don’t really believe it exists.
Second, extremist idealism leads to bad stuff. Of course, if extreme idealism were true, we should be willing to endure the bad stuff. But we also should be aware that there is a price to pay for casually accepting the idea that KurzByte is conscious.
The bad result is that we exacerbate our culture’s dismissal of flesh as an afterthought or degradation. From Plato through some of the more mortifying forms of Christianity, we already have enough impetus to despise our bodies. We really don’t need the digerati telling us that our flesh is nothing but hardware that we won’t miss once we move our software onto a new, odorless platform.
In despising flesh, we also deny the fact of death. In denying death, we also declare our complete, final and total mastery over nature. What’s the worst thing nature throws at us? Death. Who’s beaten it? We have! We’re so special.
We are special in some ways. We’re humans, dammit. But we’re also made of the earth. We go back to the earth. (Of course, we are also made of stars, if you want to feel a little better about yourself.) Flesh is the miracle that connects us to the world into which we are thrown and that lets us climb up its low peaks at least part of the way. The fact that some of the smartest people in our culture have confused representations of consciousness with consciousness itself shows just how desperately we want to escape the miracle of flesh.