How To Create A Mind Part 9

What the Libet experiments do underscore is that there is a lot of activity in our brains underlying our decisions that is not conscious. We already knew that most of what goes in the neocortex is not conscious; it should not be surprising, therefore, that our actions and decisions stem from both unconscious and conscious activity. Is this distinction important? If our decisions arise from both, should it matter if we sort out the conscious parts from the unconscious? Is it not the case that both aspects represent our brain? Are we not ultimately responsible for everything that goes on in our brains? "Yes, I shot the victim, but I'm not responsible because I wasn't paying attention" is probably a weak defense. Even though there are some narrow legal grounds on which a person may not be held responsible for his decisions, we are generally held accountable for all of the choices we make.

The observations and experiments I have cited above const.i.tute thought experiments on the issue of free will, a subject that, like the topic of consciousness, has been debated since Plato. The term "free will" itself dates back to the thirteenth century, but what exactly does it mean?

The Merriam-Webster dictionary defines it as the "freedom of humans to make choices that are not determined by prior causes or by divine intervention." You will notice that this definition is hopelessly circular: "Free will is freedom...." Setting aside the idea of divine intervention's standing in opposition to free will, there is one useful element in this definition, which is the idea of a decision's "not [being] determined by prior causes." I'll come back to that momentarily.

The Stanford Encyclopedia of Philosophy states that free will is the "capacity of rational agents to choose a course of action from among various alternatives." By this definition, a simple computer is capable of free will, so it is less helpful than the dictionary definition. states that free will is the "capacity of rational agents to choose a course of action from among various alternatives." By this definition, a simple computer is capable of free will, so it is less helpful than the dictionary definition.

Wikipedia is actually a bit better. It defines free will as "the ability of agents to make choices free from certain kinds of constraints.... The constraint of dominant concern has been...determinism." Again, it uses the circular word "free" in defining free will, but it does articulate what has been regarded as the princ.i.p.al enemy of free will: determinism determinism. In that respect the Merriam-Webster definition above is actually similar in its reference to decisions that "are not determined by prior causes."

So what do we mean by determinism? If I put "2 + 2" into a calculator and it displays "4," can I say that the calculator displayed its free will by deciding to display that "4"? No one would accept that as a demonstration of free will, because the "decision" was predetermined by the internal mechanisms of the calculator and the input. If I put in a more complex calculation, we still come to the same conclusion with regard to its lack of free will.

How about Watson when it answers a Jeopardy! Jeopardy! query? Although its deliberations are far more complex than those of the calculator, very few if any observers would ascribe free will to its decisions. No one human knows exactly how all of its programs work, but we can identify a group of people who collectively can describe all of its methods. More important, its output is determined by (1) all of its programs at the moment that the query is posed, (2) the query itself, (3) the state of its internal parameters that influence its decisions, and (4) its trillions of bytes of knowledge bases, including encyclopedias. Based on these four categories of information, its output is determined. We might speculate that presenting the same query would always get the same response, but Watson is programmed to learn from its experience, so there is the possibility that subsequent answers would be different. However, that does not contradict this a.n.a.lysis; rather, it just const.i.tutes a change in item 3, the parameters that control its decisions. query? Although its deliberations are far more complex than those of the calculator, very few if any observers would ascribe free will to its decisions. No one human knows exactly how all of its programs work, but we can identify a group of people who collectively can describe all of its methods. More important, its output is determined by (1) all of its programs at the moment that the query is posed, (2) the query itself, (3) the state of its internal parameters that influence its decisions, and (4) its trillions of bytes of knowledge bases, including encyclopedias. Based on these four categories of information, its output is determined. We might speculate that presenting the same query would always get the same response, but Watson is programmed to learn from its experience, so there is the possibility that subsequent answers would be different. However, that does not contradict this a.n.a.lysis; rather, it just const.i.tutes a change in item 3, the parameters that control its decisions.

So how exactly does a human differ from Watson, such that we ascribe free will to the human but not to the computer program? We can identify several factors. Even though Watson is a better Jeopardy! Jeopardy! player than most if not all humans, it is nonetheless not nearly as complex as a human neocortex. Watson does possess a lot of knowledge, and it does use hierarchical methods, but the complexity of its hierarchical thinking is still considerably less than that of a human. So is the difference simply one of the scale of complexity of its hierarchical thinking? There is an argument to be made that the issue does come down to this. In my discussion of the issue of consciousness I noted that my own leap of faith is that I would consider a computer that pa.s.sed a valid Turing test to be conscious. The best chatbots are not able to do that today (although they are steadily improving), so my conclusion with regard to consciousness is a matter of the level of performance of the ent.i.ty. Perhaps the same is true of my ascribing free will to it. player than most if not all humans, it is nonetheless not nearly as complex as a human neocortex. Watson does possess a lot of knowledge, and it does use hierarchical methods, but the complexity of its hierarchical thinking is still considerably less than that of a human. So is the difference simply one of the scale of complexity of its hierarchical thinking? There is an argument to be made that the issue does come down to this. In my discussion of the issue of consciousness I noted that my own leap of faith is that I would consider a computer that pa.s.sed a valid Turing test to be conscious. The best chatbots are not able to do that today (although they are steadily improving), so my conclusion with regard to consciousness is a matter of the level of performance of the ent.i.ty. Perhaps the same is true of my ascribing free will to it.

Consciousness is indeed one philosophical difference between human brains and contemporary software programs. We consider human brains to be conscious, whereas we do not-yet-attribute that to software programs. Is this the factor we are looking for that underlies free will?

A simple mind experiment would argue that consciousness is indeed a vital part of free will. Consider a situation in which someone performs an action with no awareness that she is doing it-it is carried out entirely by nonconscious activity in that person's brain. Would we regard this to be a display of free will? Most people would answer no. If the action was harmful, we would probably still hold that person responsible but look for some recent conscious acts that may have caused that person to perform actions without conscious awareness, such as taking one drink too many, or just failing to train herself adequately to consciously consider her decisions before she acted on them.

According to some commentators, the Libet experiments argued against free will by highlighting how much of our decision making is not conscious. Since there is a reasonable consensus among philosophers that free will does imply conscious decision making, it appears to be one prerequisite for free will. However, to many observers, consciousness is a necessary but not sufficient condition. If our decisions-conscious or otherwise-are predetermined before we make them, how can we say that our decisions are free? This position, which holds that free will and determinism are not compatible, is known as incompatibilism. For example, American philosopher Carl Ginet (born in 1932) argues that if events in the past, present, and future are determined, then we can be considered to have no control over them or their consequences. Our apparent decisions and actions are simply part of this predetermined sequence. To Ginet, this rules out free will.

Not everyone regards determinism as being incompatible with the concept of free will, however. The compatibilists argue, essentially, that you're free to decide what you want even though what you decide is or may be determined. Daniel Dennett, for example, argues that while the future may be determined from the state of the present, the reality is that the world is so intricately complex that we cannot possibly know what the future will bring. We can identify what he refers to as "expectations," and we are indeed free to perform acts that differ from these expectations. We should consider how our decisions and actions compare to these expectations, not to a theoretically determined future that we cannot in fact know. That, Dennett argues, is sufficient for free will.

Gazzaniga also articulates a compatibilist position: "We are personally responsible agents and are to be held accountable for our actions, even though we live in a determined world."17 A cynic might interpret this view as: You have no control over your actions, but we'll blame you anyway. A cynic might interpret this view as: You have no control over your actions, but we'll blame you anyway.

Some thinkers dismiss the idea of free will as an illusion. Scottish philosopher David Hume (17111776) described it as simply a "verbal" matter characterized by "a false sensation or seeming experience."18 German philosopher Arthur Schopenhauer (17881860) wrote that "everyone believes himself German philosopher Arthur Schopenhauer (17881860) wrote that "everyone believes himself a priori a priori to be perfectly free, even in his individual actions, and thinks that at every moment he can commence another manner of life.... But to be perfectly free, even in his individual actions, and thinks that at every moment he can commence another manner of life.... But a posteriori a posteriori, through experience, he finds to his astonishment that he is not free, but subjected to necessity, that in spite of all his resolutions and reflections he does not change his conduct, and that from the beginning of his life to the end of it, he must carry out the very character which he himself condemns."19 I would add several points here. The concept of free will-and responsibility, which is a closely aligned idea-is useful, and indeed vital, to maintaining social order, whether or not free will actually exists. Just as consciousness clearly exists as a meme, so too does free will. Attempts to prove its existence, or even to define it, may become hopelessly circular, but the reality is that almost everyone believes in the idea. Very substantial portions of our higher-level neocortex are devoted to the concept that we make free choices and are responsible for our actions. Whether in a strict philosophical sense that is true or even possible, society would be far worse off if we did not have such beliefs.

Furthermore, the world is not necessarily determined. I discussed above two perspectives on quantum mechanics, which differ with respect to the relations.h.i.+p of quantum fields to an observer. A popular interpretation of the observer-based perspective provides a role for consciousness: Particles do not resolve their quantum ambiguity until observed by a conscious observer. There is another split in the philosophy of quantum events that has a bearing on our discussion of free will, one that revolves around the question: Are quantum events determined or random?

The most common interpretation of a quantum event is that when the wave function const.i.tuting a particle "collapses," the particle's location becomes specific. Over a great many such events, there will be a predictable distribution (which is why the wave function is considered to be a probability distribution), but the resolution for each such particle undergoing a collapse of its wave function is random. The opposing interpretation is deterministic: specifically, that there is a hidden variable that we are unable to detect separately, but whose value determines the particle's position. The value or phase of the hidden variable at the moment of the wave function collapse determines the position of the particle. Most quantum physicists seem to favor the idea of a random resolution according to the probability field, but the equations for quantum mechanics do allow for the existence of such a hidden variable.

Thus the world may not be determined after all. According to the probability wave interpretation of quantum mechanics, there is a continual source of uncertainty at the most basic level of reality. However, this observation does not necessarily resolve the concerns of the incompatibilists. It is true that under this interpretation of quantum mechanics, the world is not determined, but our concept of free will extends beyond decisions and actions that are merely random. Most incompatibilists would find the concept of free will to also be incompatible with our decisions' being essentially accidental. Free will seems to imply purposeful decision making.

Dr. Wolfram proposes a way to resolve the dilemma. His book A New Kind of Science A New Kind of Science (2002) presents a comprehensive view of the idea of cellular automata and their role in every facet of our lives. A cellular automaton is a mechanism in which the value of information cells is continually recomputed as a function of the cells near it. John von Neumann created a theoretical self-replicating machine called a universal constructor that was perhaps the first cellular automaton. (2002) presents a comprehensive view of the idea of cellular automata and their role in every facet of our lives. A cellular automaton is a mechanism in which the value of information cells is continually recomputed as a function of the cells near it. John von Neumann created a theoretical self-replicating machine called a universal constructor that was perhaps the first cellular automaton.

Dr. Wolfram ill.u.s.trates his thesis with the simplest possible cellular automata, a group of cells in a one-dimensional line. At each point in time, each cell can have one of two values: black or white. The value of each cell is recomputed for each cycle. The value of a cell for the next cycle is a function of its current value as well as the value of its two adjacent neighbors. Each cellular automaton is characterized by a rule that determines how we compute whether a cell is black or white in the next cycle.

Consider the example of what Dr. Wolfram calls rule 222.

The eight possible combinations of value for the cell being recomputed and its left and right neighbors are shown in the top row. Its new value is shown in the bottom row. So, for example, if the cell is black and its two neighbors are also black, then the cell will remain black in the next generation (see the leftmost subrule of rule 222). If the cell is white, its left neighbor is white, and its right neighbor is black, then it will be changed to black in the next generation (see the subrule of rule 222 that is second from the right).

The universe for this simple cellular automaton is just one row of cells. If we start with just one black cell in the middle and show the evolution of the cells over multiple generations (where each row as we move down represents a new generation of values), the results of rule 222 look like this:

An automaton is based on a rule, and a rule defines whether the cell will be black or white based on which of the eight possible patterns exist in the current generation. Thus there are 28 = 256 possible rules. Dr. Wolfram listed all 256 possible such automata and a.s.signed each a Wolfram code from 0 to 255. Interestingly, these 256 theoretical machines have very different properties. The automata in what Dr. Wolfram calls cla.s.s I, such as rule 222, create very predictable patterns. If I were to ask what the value of the middle cell was after a trillion trillion iterations of rule 222, you could answer easily: black. = 256 possible rules. Dr. Wolfram listed all 256 possible such automata and a.s.signed each a Wolfram code from 0 to 255. Interestingly, these 256 theoretical machines have very different properties. The automata in what Dr. Wolfram calls cla.s.s I, such as rule 222, create very predictable patterns. If I were to ask what the value of the middle cell was after a trillion trillion iterations of rule 222, you could answer easily: black.

Much more interesting, however, are the cla.s.s IV automata, ill.u.s.trated by rule 110.

Multiple generations of this automaton look like this:

The interesting thing about the rule 110 automaton and cla.s.s IV automata in general is that the results are completely unpredictable. The results pa.s.s the strictest mathematical tests for randomness, yet they do not simply generate noise: There are repeating patterns, but they repeat in odd and unpredictable ways. If I were to ask you what the value of a particular cell was after a trillion trillion iterations, there would be no way to answer that question without actually running this machine through that many generations. The solution is clearly determined, because this is a very simple deterministic machine, but it is completely unpredictable without actually running the machine.

Dr. Wolfram's primary thesis is that the world is one big cla.s.s IV cellular automaton. The reason that his book is t.i.tled A New A New Kind of Science Kind of Science is because this theory contrasts with most other scientific laws. If there is a satellite orbiting Earth, we can predict where it will be five years from now without having to run through each moment of a simulated process by using the relevant laws of gravity and solve where it will be at points in time far in the future. But the future state of cla.s.s IV cellular automata cannot be predicted without simulating every step along the way. If the universe is a giant cellular automaton, as Dr. Wolfram postulates, there would be no computer big enough-since every computer would be a subset of the universe-that could run such a simulation. Therefore the future state of the universe is completely unknowable even though it is deterministic. is because this theory contrasts with most other scientific laws. If there is a satellite orbiting Earth, we can predict where it will be five years from now without having to run through each moment of a simulated process by using the relevant laws of gravity and solve where it will be at points in time far in the future. But the future state of cla.s.s IV cellular automata cannot be predicted without simulating every step along the way. If the universe is a giant cellular automaton, as Dr. Wolfram postulates, there would be no computer big enough-since every computer would be a subset of the universe-that could run such a simulation. Therefore the future state of the universe is completely unknowable even though it is deterministic.

Thus even though our decisions are determined (because our bodies and brains are part of a deterministic universe), they are nonetheless inherently unpredictable because we live in (and are part of) a cla.s.s IV automaton. We cannot predict the future of a cla.s.s IV automaton except to let the future unfold. For Dr. Wolfram, this is sufficient to allow for free will.

We don't have to look to the universe to see future events that are determined yet unpredictable. None of the scientists who have worked on Watson can predict what it will do, because the program is just too complex and varied, and its performance is based on knowledge that is far too extensive for any human to master. If we believe that humans exhibit free will, then it follows that we have to allow that future versions of Watson or Watson-like machines can exhibit it also.

My own leap of faith is that I believe that humans have free will, and while I act as if that is the case, I am hard pressed to find examples among my own decisions that ill.u.s.trate that. Consider the decision to write this book-I never made that decision. Rather, the idea of the book decided that for me. In general, I find myself captive to ideas that seem to implant themselves in my neocortex and take over. How about the decision to get married, which I made (in collaboration with one other person) thirty-six years ago? At the time, I had been following the usual program of being attracted to-and pursuing-a pretty girl. I then fell in love. Where is the free will in that?

But what about the little decisions I make every day-for example, the specific words I choose to write in my book? I start with a blank virtual sheet of paper. No one is telling me what to do. There is no editor looking over my shoulder. My choices are entirely entirely up to me. I am free- up to me. I am free-totally free-to write whatever whatever I... I...

Uh, grok grok...

Grok? Okay, I did it-I finally applied my free will. I was going to write the word "want," but I made a free decision to write something totally unexpected instead. This is perhaps the first time I've succeeded in exercising pure free will.

Or not.

It should be apparent that that was a display not of will, but rather of trying to ill.u.s.trate a point (and perhaps a weak sense of humor).

Although I share Descartes' confidence that I am conscious, I'm not so sure about free will. It is difficult to escape Schopenhauer's conclusion that "you can do what you will, but in any given moment of your life you can will will only one definite thing and absolutely nothing other than that one thing." only one definite thing and absolutely nothing other than that one thing."20 Nonetheless I will continue to act as if I have free will and to believe in it, so long as I don't have to explain why. Nonetheless I will continue to act as if I have free will and to believe in it, so long as I don't have to explain why.

Ident.i.tyA philosopher once had the following dream.First Aristotle appeared, and the philosopher said to him, "Could you give me a fifteen-minute capsule sketch of your entire philosophy?" To the philosopher's surprise, Aristotle gave him an excellent exposition in which he compressed an enormous amount of material into a mere fifteen minutes. But then the philosopher raised a certain objection which Aristotle couldn't answer. Confounded, Aristotle disappeared.Then Plato appeared. The same thing happened again, and the philosopher's objection to Plato was the same as his objection to Aristotle. Plato also couldn't answer it and disappeared.Then all the famous philosophers of history appeared one by one and our philosopher refuted every one with the same objection.After the last philosopher vanished, our philosopher said to himself, "I know I'm asleep and dreaming all this. Yet I've found a universal refutation for all philosophical systems! Tomorrow when I wake up, I will probably have forgotten it, and the world will really miss something!" With an iron effort, the philosopher forced himself to wake up, rush over to his desk, and write down his universal refutation. Then he jumped back into bed with a sigh of relief.The next morning when he awoke, he went over to the desk to see what he had written. It was, "That's what you you say." say."-Raymond Smullyan, as quoted by David Chalmers21

What I wonder about ever more than whether or not I am conscious or exercise free will is why I happen to be conscious of the experiences and decisions of this one particular person who writes books, enjoys hiking and biking, takes nutritional supplements, and so on. An obvious answer would be, "Because that's who you are."

That exchange is probably no more tautological than my answers above to questions about consciousness and free will. But actually I do have a better answer for why my consciousness is a.s.sociated with this particular person: It is because that is who I created myself to be.

A common aphorism is, "You are what you eat." It is even more true to say, "You are what you think." As we have discussed, all of the hierarchical structures in my neocortex that define my personality, skills, and knowledge are the result of my own thoughts and experiences. The people I choose to interact with and the ideas and projects I choose to engage in are all primary determinants of who I become. For that matter, what I eat also reflects the decisions made by my neocortex. Accepting the positive side of the free will duality for the moment, it is my own decisions that result in who I am.

Regardless of how we came to be who we are, each of us has the desire for our ident.i.ty to persist. If you didn't have the will to survive, you wouldn't be here reading this book. Every creature has that goal-it is the princ.i.p.al determinant of evolution. The issue of ident.i.ty is perhaps even harder to define than consciousness or free will, but is arguably more important. After all, we need to know what we are if we seek to preserve our existence.

Consider this thought experiment: You are in the future with technologies more advanced than today's. While you are sleeping, some group scans your brain and picks up every salient detail. Perhaps they do this with blood cellsized scanning machines traveling in the capillaries of your brain or with some other suitable noninvasive technology, but they have all of the information about your brain at a particular point in time. They also pick up and record any bodily details that might reflect on your state of mind, such as the endocrine system. They instantiate this "mind file" in a nonbiological body that looks and moves like you and has the requisite subtlety and suppleness to pa.s.s for you. In the morning you are informed about this transfer and you watch (perhaps without being noticed) your mind clone, whom we'll call You 2. You 2 is talking about his or her life as if s/he were you, and relating how s/he discovered that very morning that s/he had been given a much more durable new version 2.0 body. "Hey, I kind of like this new body!" s/he exclaims.

The first question to consider is: Is You 2 conscious? Well, s/he certainly seems to be. S/he pa.s.ses the test I articulated earlier, in that s/he has the subtle cues of being a feeling, conscious person. If you are conscious, then so too is You 2.

So if you were to, uh, disappear, no one would notice. You 2 would go around claiming to be you. All of your friends and loved ones would be content with the situation and perhaps pleased that you now have a more durable body and mental substrate than you used to have. Perhaps your more philosophically minded friends would express concerns, but for the most part, everybody would be happy, including you, or at least the person who is convincingly claiming to be you.

So we don't need your old body and brain anymore, right? Okay if we dispose of it?

You're probably not going to go along with this. I indicated that the scan was noninvasive, so you are still around and still conscious. Moreover your sense of ident.i.ty is still with you, not with You 2, even though You 2 thinks s/he is a continuation of you. You 2 might not even be aware that you exist or ever existed. In fact you would not be aware of the existence of You 2 either, if we hadn't told you about it.

Our conclusion? You 2 is conscious but is a different person than you-You 2 has a different ident.i.ty. S/he is extremely similar, much more so than a mere genetic clone, because s/he also shares all of your neocortical patterns and connections. Or I should say s/he shared those patterns at the moment s/he was created. At that point, the two of you started to go your own ways, neocortically speaking. You are still around. You are not having the same experiences as You 2. Bottom line: You 2 is not you.

Okay, so far so good. Now consider another thought experiment-one that is, I believe, more realistic in terms of what the future will bring. You undergo a procedure to replace a very small part of your brain with a nonbiological unit. You're convinced that it's safe, and there are reports of various benefits.

This is not so far-fetched, as it is done routinely for people with neurological and sensory impairments, such as the neural implant for Parkinson's disease and cochlear implants for the deaf. In these cases the computerized device is placed inside the body but outside the brain yet connected into the brain (or in the case of the cochlear implants, to the auditory nerve). In my view the fact that the actual computer is physically placed outside the actual brain is not philosophically significant: We are effectively augmenting the brain and replacing with a computerized device those of its functions that no longer work properly. In the 2030s, when intelligent computerized devices will be the size of blood cells (and keep in mind that white blood cells are sufficiently intelligent to recognize and combat pathogens), we will introduce them noninvasively, no surgery required.

Returning to our future scenario, you have the procedure, and as promised, it works just fine-certain of your capabilities have improved. (You have better memory, perhaps.) So are you still you? Your friends certainly think so. You think so. There is no good argument that you're suddenly a different person. Obviously, you underwent the procedure in order to effect a change in something, but you are still the same you. Your ident.i.ty hasn't changed. Someone else's consciousness didn't suddenly take over your body.

Okay, so, encouraged by these results, you now decide to have another procedure, this time involving a different region of the brain. The result is the same: You experience some improvement in capability, but you're still you.

It should be apparent where I am going with this. You keep opting for additional procedures, your confidence in the process only increasing, until eventually you've changed every part of your brain. Each time the procedure was carefully done to preserve all of your neocortical patterns and connections so that you have not lost any of your personality, skills, or memories. There was never a you and a You 2; there was only you. No one, including you, ever notices you ceasing to exist. Indeed-there you are.

Our conclusion: You still exist. There's no dilemma here. Everything is fine.

Except for this: You, after the gradual replacement process, are entirely equivalent to You 2 in the prior thought experiment (which I will call the scan-and-instantiate scenario). You, after the gradual replacement scenario, have all of the neocortical patterns and connections that you had originally, only in a nonbiological substrate, which is also true of You 2 in the scan-and-instantiate scenario. You, after the gradual replacement scenario, have some additional capabilities and greater durability than you did before the process, but this is likewise true of You 2 in the scan-and-instantiate process. You, after the gradual replacement process, are entirely equivalent to You 2 in the prior thought experiment (which I will call the scan-and-instantiate scenario). You, after the gradual replacement scenario, have all of the neocortical patterns and connections that you had originally, only in a nonbiological substrate, which is also true of You 2 in the scan-and-instantiate scenario. You, after the gradual replacement scenario, have some additional capabilities and greater durability than you did before the process, but this is likewise true of You 2 in the scan-and-instantiate process.

But we concluded that You 2 is not not you. And if you, after the gradual replacement process, are entirely equivalent to You 2 after the scan-and-instantiate process, then you after the gradual replacement process must also not be you. you. And if you, after the gradual replacement process, are entirely equivalent to You 2 after the scan-and-instantiate process, then you after the gradual replacement process must also not be you.

That, however, contradicts our earlier conclusion. The gradual replacement process consists of multiple steps. Each of those steps appeared to preserve ident.i.ty, just as we conclude today that a Parkinson's patient has the same ident.i.ty after having had a neural implant installed.22 It is just this sort of philosophical dilemma that leads some people to conclude that these replacement scenarios will never happen (even though they are already taking place). But consider this: We naturally undergo a gradual replacement process throughout our lives. Most of our cells in our body are continuously being replaced. (You just replaced 100 million of them in the course of reading the last sentence.) Cells in the inner lining of the small intestine turn over in about a week, as does the stomach's protective lining. The life span of white blood cells ranges from a few days to a few months, depending on the type. Platelets last about nine days.

Neurons persist, but their organelles and their const.i.tuent molecules turn over within a month.23 The half-life of a neuron microtubule is about ten minutes; the actin filaments in the dendrites last about forty seconds; the proteins that provide energy to the synapses are replaced every hour; the NMDA receptors in synapses are relatively long-lived at five days. The half-life of a neuron microtubule is about ten minutes; the actin filaments in the dendrites last about forty seconds; the proteins that provide energy to the synapses are replaced every hour; the NMDA receptors in synapses are relatively long-lived at five days.

So you are completely replaced in a matter of months, which is comparable to the gradual replacement scenario I describe above. Are you the same person you were a few months ago? Certainly there are some differences. Perhaps you learned a few things. But you a.s.sume that your ident.i.ty persists, that you are not continually destroyed and re-created.

Consider a river, like the one that flows past my office. As I look out now at what people call the Charles River, is it the same river that I saw yesterday? Let's first reflect on what a river is. The dictionary defines it is "a large natural stream of flowing water." By that definition, the river I'm looking at is a completely different one than it was yesterday. Every one of its water molecules has changed, a process that happens very quickly. Greek philosopher Diogenes Laertius wrote in the third century AD that "you cannot step into the same river twice."

But that is not how we generally regard rivers. People like to look at them because they are symbols of continuity and stability. By the common view, the Charles River that I looked at yesterday is the same river I see today. Our lives are much the same. Fundamentally we are not the stuff that makes up our bodies and brains. These particles essentially flow through us in the same way that water molecules flow through a river. We are a pattern that changes slowly but has stability and continuity, even though the stuff const.i.tuting the pattern changes quickly.

The gradual introduction of nonbiological systems into our bodies and brains will be just another example of the continual turnover of parts that compose us. It will not alter the continuity of our ident.i.ty any more than the natural replacement of our biological cells does. We have already largely outsourced our historical, intellectual, social, and personal memories to our devices and the cloud. The devices we interact with to access these memories may not yet be inside our bodies and brains, but as they become smaller and smaller (and we are shrinking technology at a rate of about a hundred in 3-D volume per decade), they will make their way there. In any event, it will be a useful place to put them-we won't lose them that way. If people do opt out of placing microscopic devices inside their bodies, that will be fine, as there will be other ways to access the pervasive cloud intelligence.

But we come back to the dilemma I introduced earlier. You, after a period of gradual replacement, are equivalent to You 2 in the scan-and-instantiate scenario, but we decided that You 2 in that scenario does not have the same ident.i.ty as you. So where does that leave us?

It leaves us with an appreciation of a capability that nonbiological systems have that biological systems do not: the ability to be copied, backed up, and re-created. We do that routinely with our devices. When we get a new smartphone, we copy over all of our files, so it has much the same personality, skills, and memories that the old smartphone did. Perhaps it also has some new capabilities, but the contents of the old phone are still with us. Similarly, a program such as Watson is certainly backed up. If the Watson hardware were destroyed tomorrow, Watson would easily be re-created from its backup files stored in the cloud.

This represents a capability in the nonbiological world that does not exist in the biological world. It is an advantage, not a limitation, which is one reason why we are so eager today to continue uploading our memories to the cloud. We will certainly continue in this direction, as nonbiological systems attain more and more of the capabilities of our biological brains.

My resolution of the dilemma is this: It is not true that You 2 is not you-it is is you. It is just that there are now two of you. That's not so bad-if you think you are a good thing, then two of you is even better. you. It is just that there are now two of you. That's not so bad-if you think you are a good thing, then two of you is even better.

What I believe will actually happen is that we will continue on the path of the gradual replacement and augmentation scenario until ultimately most of our thinking will be in the cloud. My leap of faith on ident.i.ty is that ident.i.ty is preserved through continuity of the pattern of information that makes us us. Continuity does allow for continual change, so whereas I am somewhat different than I was yesterday, I nonetheless have the same ident.i.ty. However, the continuity of the pattern that const.i.tutes my ident.i.ty is not substrate-dependent. Biological substrates are wonderful-they have gotten us very far-but we are creating a more capable and durable substrate for very good reasons.

CHAPTER 10

THE LAW OF ACCELERATING RETURNS APPLIED TO THE BRAIN

And though man should remain, in some respects, the higher creature, is not this in accordance with the practice of nature, which allows superiority in some things to animals which have, on the whole, been long surpa.s.sed? Has she not allowed the ant and the bee to retain superiority over man in the organization of their communities and social arrangements, the bird in traversing the air, the fish in swimming, the horse in strength and fleetness, and the dog in self-sacrifice?-Samuel Butler, 1871 There was a time, when the earth was to all appearance utterly dest.i.tute both of animal and vegetable life, and when according to the opinion of our best philosophers it was simply a hot round ball with a crust gradually cooling. Now if a human being had existed while the earth was in this state and had been allowed to see it as though it were some other world with which he had no concern, and if at the same time he were entirely ignorant of all physical science, would he not have p.r.o.nounced it impossible that creatures possessed of anything like consciousness should be evolved from the seeming cinder which he was beholding? Would he not have denied that it contained any potentiality of consciousness? Yet in the course of time consciousness came. Is it not possible then that there may be even yet new channels dug out for consciousness, though we can detect no signs of them at present?-Samuel Butler, 1871 When we reflect upon the manifold phases of life and consciousness which have been evolved already, it would be rash to say that no others can be developed, and that animal life is the end of all things. There was a time when fire was the end of all things: another when rocks and water were so.-Samuel Butler, 1871 There is no security against the ultimate development of mechanical consciousness, in the fact of machines possessing little consciousness now. A mollusk has not much consciousness. Reflect upon the extraordinary advance which machines have made during the last few hundred years, and note how slowly the animal and vegetable kingdoms are advancing. The more highly organized machines are creatures not so much of yesterday, as of the last five minutes, so to speak, in comparison with past time. a.s.sume for the sake of argument that conscious beings have existed for some twenty million years: see what strides machines have made in the last thousand! May not the world last twenty million years longer? If so, what will they not in the end become?-Samuel Butler, 1871

My core thesis, which I call the law of accelerating returns (LOAR), is that fundamental measures of information technology follow predictable and exponential trajectories, belying the conventional wisdom that "you can't predict the future." There are still many things-which project, company, or technical standard will prevail in the marketplace, when peace will come to the Middle East-that remain unknowable, but the underlying price/performance and capacity of information has nonetheless proven to be remarkably predictable. Surprisingly, these trends are unperturbed by conditions such as war or peace and prosperity or recession.

A primary reason that evolution created brains was to predict the future. As one of our ancestors walked through the savannas thousands of years ago, she might have noticed that an animal was progressing toward a route that she was taking. She would predict that if she stayed on course, their paths would intersect. Based on this, she decided to head in another direction, and her foresight proved valuable to survival.

But such built-in predictors of the future are linear, not exponential, a quality that stems from the linear organization of the neocortex. Recall that the neocortex is constantly making predictions-what letter and word we will see next, whom we expect to see as we round the corner, and so on. The neocortex is organized with linear sequences of steps in each pattern, which means that exponential thinking does not come naturally to us. The cerebellum also uses linear predictions. When it helps us to catch a fly ball it is making a linear prediction about where the ball will be in our visual field of view and where our gloved hand should be in our visual field of view to catch it.

As I have pointed out, there is a dramatic difference between linear and exponential progressions (forty steps linearly is forty, but exponentially is a trillion), which accounts for why my predictions stemming from the law of accelerating returns seem surprising to many observers at first. We have to train ourselves to think exponentially. When it comes to information technologies, it is the right way to think.

The quintessential example of the law of accelerating returns is the perfectly smooth, doubly exponential growth of the price/performance of computation, which has held steady for 110 years through two world wars, the Great Depression, the Cold War, the collapse of the Soviet Union, the reemergence of China, the recent financial crisis, and all of the other notable events of the late nineteenth, twentieth, and early twenty-first centuries. Some people refer to this phenomenon as "Moore's law," but that is a misconception. Moore's law-which states that you can place twice as many components on an integrated circuit every two years, and they run faster because they are smaller-is just one paradigm among many. It was in fact the fifth, not the first, paradigm to bring exponential growth to the price/performance of computing.

The exponential rise of computation started with the 1890 U.S. census (the first to be automated) using the first paradigm of electromechanical calculation, decades before Gordon Moore was even born. In The Singularity Is Near The Singularity Is Near I provide this graph through 2002, and here I update it through 2009 (see the graph on I provide this graph through 2002, and here I update it through 2009 (see the graph on page 257 page 257 t.i.tled "Exponential Growth of Computing for 110 Years"). The smoothly predictable trajectory has continued, even through the recent economic downturn. t.i.tled "Exponential Growth of Computing for 110 Years"). The smoothly predictable trajectory has continued, even through the recent economic downturn.

Computation is the most important example of the law of accelerating returns, because of the amount of data we have for it, the ubiquity of computation, and its key role in ultimately revolutionizing everything we care about. But it is far from the only example. Once a technology becomes an information technology, it becomes subject to the LOAR.

Biomedicine is becoming the most significant recent area of technology and industry to be transformed in this way. Progress in medicine has historically been based on accidental discoveries, so progress during the earlier era was linear, not exponential. This has nevertheless been beneficial: Life expectancy has grown from twenty-three years as of a thousand years ago, to thirty-seven years as of two hundred years ago, to close to eighty years today. With the gathering of the software of life-the genome-medicine and human biology have become an information technology. The human genome project itself was perfectly exponential, with the amount of genetic data doubling and the cost per base pair coming down by half each year since the project was initiated in 1990.3 (All the graphs in this chapter have been updated since (All the graphs in this chapter have been updated since The Singularity Is Near The Singularity Is Near was published.) was published.)

The cost of sequencing a human-sized genome.1

The amount of genetic data sequenced in the world each year.2

We now have the ability to design biomedical interventions on computers and to test them on biological simulators, the scale and precision of which are also doubling every year. We can also update our own obsolete software: RNA interference can turn genes off, and new forms of gene therapy can add new genes, not just to a newborn but to a mature individual. The advance of genetic technologies also affects the brain reverse-engineering project, in that one important aspect of it is understanding how genes control brain functions such as creating new connections to reflect recently added cortical knowledge. There are many other manifestations of this integration of biology and information technology, as we move beyond genome sequencing to genome synthesizing.

Another information technology that has seen smooth exponential growth is our ability to communicate with one another and transmit vast repositories of human knowledge. There are many ways to measure this phenomenon. Cooper's law, which states that the total bit capacity of wireless communications in a given amount of radio spectrum doubles every thirty months, has held true from the time Guglielmo Marconi used the wireless telegraph for Morse code transmissions in 1897 to today's 4G communications technologies.4 According to Cooper's law, the amount of information that can be transmitted over a given amount of radio spectrum has been doubling every two and a half years for more than a century. Another example is the number of bits per second transmitted on the Internet, which is doubling every one and a quarter years. According to Cooper's law, the amount of information that can be transmitted over a given amount of radio spectrum has been doubling every two and a half years for more than a century. Another example is the number of bits per second transmitted on the Internet, which is doubling every one and a quarter years.5 The reason I became interested in trying to predict certain aspects of technology is that I realized about thirty years ago that the key to becoming successful as an inventor (a profession I adopted when I was five years old) was timing. Most inventions and inventors fail not because the gadgets themselves don't work, but because their timing is wrong, appearing either before all of the enabling factors are in place or too late, having missed the window of opportunity.

The international (country-to-country) bandwidth dedicated to the Internet for the world.6