by Ray Kurzweil
You might also point out that the premise is absurd. Just changing the mechanical linkages in a mechanical typewriter could not possibly enable it to convincingly answer questions in Chinese (not to mention the fact that we can’t fit the thousands of Chinese-character symbols on the keys of a typewriter).
Yes, that’s a valid objection, as well. The only difference between my Chinese Room conception and the several proposed by Searle is that it is patently obvious in my conception that it couldn’t possibly work and is by its very nature absurd. That may not be quite as apparent to many readers or listeners with regard to the Searle Chinese Rooms. However, it is equally the case.
And yet we can make my conception work, just as we can make Searle’s conceptions work. All you have to do is to make the typewriter linkages as complex as a human brain. And that’s theoretically (if not practically) possible. But the phrase “typewriter linkages” does not suggest such vast complexity. The same is true of Searle’s description of a person manipulating slips of paper or following a book of rules or a computer program. These are all equally misleading conceptions.
Searle writes: “Actual human brains cause consciousness by a series of specific neurobiological processes in the brain.” However, he has yet to provide any basis for such a startling view. To illuminate Searle’s perspective, I quote from a letter he sent me:
It may turn out that rather simple organisms like termites or snails are conscious. . . . The essential thing is to recognize that consciousness is a biological process like digestion, lactation, photosynthesis, or mitosis, and you should look for its specific biology as you look for the specific biology of these other processes.38
I replied:
Yes, it is true that consciousness emerges from the biological process(es) of the brain and body, but there is at least one difference. If I ask the question, “does a particular entity emit carbon dioxide,” I can answer that question through clear objective measurement. If I ask the question, “is this entity conscious,” I may be able to provide inferential arguments—possibly strong and convincing ones—but not clear objective measurement.
With regard to the snail, I wrote:
Now when you say that a snail may be conscious, I think what you are saying is the following: that we may discover a certain neurophysiological basis for consciousness (call it “x”) in humans such that when this basis was present humans were conscious, and when it was not present humans were not conscious. So we would presumably have an objectively measurable basis for consciousness. And then if we found that in a snail, we could conclude that it was conscious. But this inferential conclusion is just a strong suggestion, it is not a proof of subjective experience on the snail’s part. It may be that humans are conscious because they have “x” as well as some other quality that essentially all humans share, call this “y.” The “y” may have to do with a human’s level of complexity or something having to do with the way we are organized, or with the quantum properties of our microtubules (although this may be part of “x”), or something else entirely. The snail has “x” but doesn’t have “y” and so it may not be conscious.
How would one settle such an argument? You obviously can’t ask the snail. Even if we could imagine a way to pose the question, and it answered yes, that still wouldn’t prove that it was conscious. You can’t tell from its fairly simple and more-or-less predictable behavior. Pointing out that it has “x” may be a good argument, and many people may be convinced by it. But it’s just an argument—not a direct measurement of the snail’s subjective experience. Once again, objective measurement is incompatible with the very concept of subjective experience.
Many such arguments are taking place today—though not so much about snails as about higher-level animals. It is apparent to me that dogs and cats are conscious (and Searle has said that he acknowledges this as well). But not all humans accept this. I can imagine scientific ways of strengthening the argument by pointing out many similarities between these animals and humans, but again these are just arguments, not scientific proof.
Searle expects to find some clear biological “cause” of consciousness, and he seems unable to acknowledge that either understanding or consciousness may emerge from an overall pattern of activity. Other philosophers, such as Daniel Dennett, have articulated such “pattern emergent” theories of consciousness. But whether it is “caused” by a specific biological process or by a pattern of activity, Searle provides no foundation for how we would measure or detect consciousness. Finding a neurological correlate of consciousness in humans does not prove that consciousness is necessarily present in other entities with the same correlate, nor does it prove that the absence of such a correlate indicates the absence of consciousness. Such inferential arguments necessarily stop short of direct measurement. In this way, consciousness differs from objectively measurable processes such as lactation and photosynthesis.
As I discussed in chapter 4, we have discovered a biological feature unique to humans and a few other primates: the spindle cells. And these cells with their deep branching structures do appear to be heavily involved with our conscious responses, especially emotional ones. Is the spindle cell structure the neurophysiological basis “x” for human consciousness? What sort of experiment could possibly prove that? Cats and dogs don’t have spindle cells. Does that prove that they have no conscious experience?
Searle writes: “It is out of the question, for purely neurobiological reasons, to suppose that the chair or the computer is conscious.” I agree that chairs don’t seem to be conscious, but as for computers of the future that have the same complexity, depth, subtlety, and capabilities as humans, I don’t think we can rule out this possibility. Searle just assumes that they are not, and that it is “out of the question” to suppose otherwise. There is really nothing more of a substantive nature to Searle’s “arguments” than this tautology.
Now, part of the appeal of Searle’s stance against the possibility of a computer’s being conscious is that the computers we know today just don’t seem to be conscious. Their behavior is brittle and formulaic, even if they are occasionally unpredictable. But as I pointed out above, computers today are on the order of one million times simpler than the human brain, which is at least one reason they don’t share all of the endearing qualities of human thought. But that disparity is rapidly shrinking and will ultimately reverse itself in a couple of decades. The early twenty-first-century machines I am talking about in this book will appear and act very differently than the relatively simple computers of today.
Searle articulates the view that nonbiological entities are capable of only manipulating logical symbols and he appears to be unaware of other paradigms. It is true that manipulating symbols is largely how rule-based expert systems and game-playing programs work. But the current trend is in a different direction, toward self-organizing chaotic systems that employ biologically inspired methods, including processes derived directly from the reverse engineering of the hundreds of neuron clusters we call the human brain.
Searle acknowledges that biological neurons are machines—indeed, that the entire brain is a machine. As I discussed in chapter 4, we have already recreated in an extremely detailed way the “causal powers” of individual neurons as well as those of substantial neuron clusters. There is no conceptual barrier to scaling these efforts up to the entire human brain.
The Criticism from the Rich-Poor Divide
Another concern expressed by Jaron Lanier and others is the “terrifying” possibility that through these technologies the rich may gain certain advantages and opportunities to which the rest of humankind does not have access.39 Such inequality, of course, would be nothing new, but with regard to this issue the law of accelerating returns has an important and beneficial impact. Because of the ongoing exponential growth of price-performance, all of these technologies quickly become so inexpensive as to become almost free.
Look at the extraordinary amount of high-quality information available at no cos
t on the Web today that did not exist at all just a few years ago. And if one wants to point out that only a fraction of the world today has Web access, keep in mind that the explosion of the Web is still in its infancy, and access is growing exponentially. Even in the poorest countries of Africa, Web access is expanding rapidly.
Each example of information technology starts out with early-adoption versions that do not work very well and that are unaffordable except by the elite. Subsequently the technology works a bit better and becomes merely expensive. Then it works quite well and becomes inexpensive. Finally it works extremely well and is almost free. The cell phone, for example, is somewhere between these last two stages. Consider that a decade ago if a character in a movie took out a portable telephone, this was an indication that this person must be very wealthy, powerful, or both. Yet there are societies around the world in which the majority of the population were farming with their hands two decades ago and now have thriving information-based economies with widespread use of cell phones (for example, Asian societies, including rural areas of China). This lag from very expensive early adopters to very inexpensive, ubiquitous adoption now takes about a decade. But in keeping with the doubling of the paradigm-shift rate each decade, this lag will be only five years a decade from now. In twenty years, the lag will be only two to three years (see chapter 2).
The rich-poor divide remains a critical issue, and at each point in time there is more that can and should be done. It is tragic, for example, that the developed nations were not more proactive in sharing AIDS drugs with poor countries in Africa and elsewhere, with millions of lives lost as a result. But the exponential improvement in the price-performance of information technologies is rapidly mitigating this divide. Drugs are essentially an information technology, and we see the same doubling of price-performance each year as we do with other forms of information technology such as computers, communications, and DNA base-pair sequencing. AIDS drugs started out not working very well and costing tens of thousands of dollars per patient per year. Today these drugs work reasonably well and are approaching one hundred dollars per patient per year in poor countries such as those in Africa.
In chapter 2 I cited the World Bank report for 2004 of higher economic growth in the developing world (over 6 percent) compared to the world average (of 4 percent), and an overall reduction in poverty (for example, a reduction of 43 percent in extreme poverty in the East Asian and Pacific region since 1990). Moreover, economist Xavier Sala-i-Martin examined eight measures of global inequality among individuals, and found that all were declining over the past quarter century.40
The Criticism from the Likelihood of Government Regulation
These guys talking here act as though the government is not part of their lives. They may wish it weren’t, but it is. As we approach the issues they debated here today, they had better believe that those issues will be debated by the whole country. The majority of Americans will not simply sit still while some elite strips off their personalities and uploads themselves into their cyberspace paradise. They will have something to say about that. There will be vehement debate about that in this country.
—LEON FUERTH, FORMER NATIONAL SECURITY ADVISER TO VICE PRESIDENT
AL GORE, AT THE 2002 FORESIGHT CONFERENCE
Human life without death would be something other than human; consciousness of mortality gives rise to our deepest longings and greatest accomplishments.
—LEON KASS, CHAIR OF THE PRESIDENTIAL COMMISSION ON BIOETHICS, 2003
The criticism concerning governmental control is that regulation will slow down and stop the acceleration of technology. Although regulation is a vital issue, it has actually had no measurable effect on the trends discussed in this book, which have occurred with extensive regulation in place. Short of a worldwide totalitarian state, the economic and other forces underlying technical progress will only grow with ongoing advances.
Consider the issue of stem-cell research, which has been especially controversial, and for which the U.S. government is restricting its funding. Stem-cell research is only one of numerous ideas concerned with controlling and influencing the information processes underlying biology that are being pursued as part of the biotechnology revolution. Even within the field of cell therapies the controversy over embryonic stem-cell research has served only to accelerate other ways of accomplishing the same goal. For example, transdifferentiation (converting one type of cell such as a skin cell into other types of cells) has moved ahead quickly.
As I reported in chapter 5, scientists have recently demonstrated the ability to reprogram skin cells into several other cell types. This approach represents the holy grail of cell therapies in that it promises an unlimited supply of differentiated cells with the patient’s own DNA. It also allows cells to be selected without DNA errors and will ultimately be able to provide extended telomere strings (to make the cells more youthful). Even embryonic stem-cell research itself has moved ahead, for example, with projects like Harvard’s major new research center and California’s successful three-billion-dollar bond initiative to support such work.
Although the restrictions on stem-cell research are unfortunate, it is hard to say that cell-therapy research, let alone the broad field of biotechnology, has been affected to a significant degree.
Some governmental restrictions reflect the perspective of fundamentalist humanism, which I addressed in the previous chapter. For example, the Council of Europe proclaimed that “human rights imply the right to inherit a genetic pattern that has not been artificially changed.”41 Perhaps the most interesting aspect of the council’s edict is its posing a restriction as a right. In the same spirit, I assume the council would advocate the human right not to be cured from natural disease by unnatural means, just as activists “protected” starving African nations from the indignity of consuming bioengineered crops.42
Ultimately the benefits of technical progress overwhelm such reflexive antitechnology sentiments. The majority of crops in the United States are already GMOs, while Asian nations are aggressively adopting the technology to feed their large populations, and even Europe is now beginning to approve GMO foods. The issue is important because unnecessary restrictions, although temporary, can result in exacerbated suffering of millions of people. But technical progress is advancing on thousands of fronts, fueled by irresistible economic gains and profound improvements in human health and well-being.
Leon Fuerth’s observation quoted above reveals an inherent misconception about information technologies. Information technologies are not available only to an elite. As discussed, desirable information technologies rapidly become ubiquitous and almost free. It is only when they don’t work very well (that is, in an early stage of development) that they are expensive and restricted to an elite.
Early in the second decade of this century, the Web will provide full immersion visual-auditory virtual reality with images written directly to our retinas from our eyeglasses and lenses and very high-bandwidth wireless Internet access woven in our clothing. These capabilities will not be restricted just to the privileged. Just like cell phones, by the time they work well they will be everywhere.
In the 2020s we will routinely have nanobots in our bloodstream keeping us healthy and augmenting our mental capabilities. By the time these work well they will be inexpensive and widely used. As I discussed above, reducing the lag between early and late adoption of information technologies will itself accelerate from the current ten-year period to only a couple of years two decades from now. Once nonbiological intelligence gets a foothold in our brains, it will at least double in capability each year, as is the nature of information technology. Thus it will not take long for the nonbiological portion of our intelligence to predominate. This will not be a luxury reserved for the rich, any more than search engines are today. And to the extent that there will be a debate about the desirability of such augmentation, it’s easy to predict who will win, since those with enhanced intelligence will be far better debaters.<
br />
The Unbearable Slowness of Social Institutions. MIT senior research scientist Joel Cutcher-Gershenfeld writes: “Just looking back over the course of the past century and a half, there have been a succession of political regimes where each was the solution to an earlier dilemma, but created new dilemmas in the subsequent era. For example, Tammany Hall and the political patron model were a vast improvement over the dominant system based on landed gentry—many more people were included in the political process. Yet, problems emerged with patronage, which led to the civil service model—a strong solution to the preceding problem by introducing the meritocracy. Then, of course, civil service became the barrier to innovation and we move to reinventing government. And the story continues.”43 Gershenfeld is pointing out that social institutions even when innovative in their day become “a drag on innovation.”
First I would point out that the conservatism of social institutions is not a new phenomenon. It is part of the evolutionary process of innovation, and the law of accelerating returns has always operated in this context. Second, innovation has a way of working around the limits imposed by institutions. The advent of decentralized technology empowers the individual to bypass all kinds of restrictions, and does represent a primary means for social change to accelerate. As one of many examples, the entire thicket of communications regulations is in the process of being bypassed by emerging point-to-point techniques such as voice over Internet protocol (VOIP).
Virtual reality will represent another means of hastening social change. People will ultimately be able to have relationships and engage in activities in immersive and highly realistic virtual-reality environments that they would not be able or willing to do in real reality.
As technology becomes more sophisticated it increasingly takes on traditional human capabilities and requires less adaptation. You had to be technically adept to use early personal computers, whereas using computerized systems today, such as cell phones, music players, and Web browsers, requires much less technical ability. In the second decade of this century, we will routinely be interacting with virtual humans that, although not yet Turing-test capable, will have sufficient natural language understanding to act as our personal assistants for a wide range of tasks.