Human Nature
Podcast # 4: Artificial Intelligence
Albert A. Anderson
Copyright 2006
I. The Value Question
A. Behaviorism, Determinism, and Materialism
When I wrote “Minds and machines: A Modest Proposal,” I had three targets in mind: (1) behaviorism, (2) determinism and (3) materialism. I was especially concerned about the implication of those ideas for moral and political values.
A. Behaviorism
Searle’s Chinese Room is partly designed to question behaviorism — the view that there is no more to mind than its effects, its behavior. John Watson and B. F. Skinner embraced behaviorism and set the stage for many versions of that approach in the 20th century. Searle’s central point is that meaning is essential for human thinking, and meaning is a function of first-person language and immediate experience. The central difference between syntax and semantics is that syntactical structures can be instilled from the outside and observed from the outside. But semantics cannot be acquired from the outside. It is the essential element not only in thinking but also in acting and choosing. Philosophers as different as Socrates, Kant, Marx, William James, Gandhi, and Sartre, all agree on one central point: to be human is to choose and to act — to shape our own destiny. For that to be possible, we must be free agents. This leads to the problem of determinism.
B. Determinism.
Darwin’s theory of evolution, at least the way it is often presented, seems to undermine and threaten human freedom.
· Perhaps the best way to explain determinism in this context is by appeal to the 18th century French philosopher Pierre Laplace. He was steeped in Newtonian physics and was convinced that Newton’s laws of motion provided a full explanation of natural phenomena. This led him to claim that an ideal observer who knew the position of every particle of matter and knew the laws governing all their movements, that observer could predict as well as reconstruct all events in the universe. In other words, the laws of physics determine everything. Genuine freedom (free choice) has no place in such a universe. If the laws of chemistry are built upon the laws of physics, chemical laws are is equally determined. If the laws of biology are built upon chemistry, they are equally determined. If the laws of psychology are built on biology, they are equally determined. And so on throughout all of the natural and social sciences. No matter what we do, no matter how meaningless this might make what we falsely think are our decisions, there is no place for free will in such a system. Replacing human beings by machines (or any other products of evolution that are more fit to survive than we are) would be inevitable.
· But this leaves us in a classical dilemma: Either (a) we abandon freedom and the moral choices that depend on free choice, or (b) we abandon the causal laws of physics, chemistry, and biology and open the way for any unscientific or anti-scientific belief that can be imagined. Either way we seem to lose something a reasonable person cannot give up. Can we resolve this dilemma?
· This view of nature stands in direct opposition to the position taken by a philosopher who started out in physics and who was an older contemporary of Pierre LaPlace: Immanuel Kant. Kant was well aware of the dilemma posed by scientific determinism. For that reason he insisted on separating two fundamentally different realms to which we have access:
“There are two possible points of view from which we can see ourselves and recognize the laws governing the use of our powers and thus all of our actions. First, insofar as we belong to the world of sensation, we see ourselves subject to the laws of nature (of heteronomy). Second, we see ourselves belonging to the intelligible world, subject to laws that are independent of nature, not empirical but grounded in reason alone. As rational beings, and consequently as belonging to the intelligible world, we can never think of the causality governing our own will except as pertaining to the idea of freedom” (Immanuel Kant, Foundations of Ethics, Agora Publications, 1995, p. 66 [452]).
· Kant refused to give up either the scientific account of the empirical world of material things or the world of the mind that is subject to a different set of laws — the laws of freedom. His distinction between (a) heteronomy and (b) autonomy lies at the center of his attempt to avoid the horns of the dilemma that emerges from the modern scientific worldview.
o Heteronomy is derived form two Greek words: hetero (other) and nomos (law). The world of nature outside of human beings is ruled by laws that are alien to us — they are imposed without concern for our will — we have no choice in the matter. The law of gravity or the speed of light from physics and natural selection or the survival of the fittest from biology are heteronomous.
o Autonomy comes from the Greek words auto (self) and nomos (law). According to Kant, causality in this realm governs our will — or decisions, choices, and actions. To this, he says, “is tied the universal principle of morality” (p. 66). All actions by what Kant calls “rational beings” are grounded in autonomy. A rational being is any being that is able to think of laws that guide free choice and choose to follow them. This is also how Kant defines what it means to be a person.
· Persons and Things
o Kant would have no difficulty in explaining the crucial difference between computers and human beings. Computers are things, totally directed and controlled by heteronomy. Human beings are persons, guided by autonomy.
o The contemporary struggle to explain why persons can and should control the world is quickly and easily resolved in Kantian terms by showing the superiority of any being that directs and controls itself and a thing that is totally determined by outside laws and forces.
o Those who wish to distinguish human beings from digital computers frequently employ two different strategies.
§ In the first strategy they point out that human beings have feelings, emotions, passions, and desires, and machines lack those phenomena. But for Kant, this strategy is doomed to fail, because it locates human uniqueness in a realm that belongs not to the autonomous realm of the mind but to the heteronomous world of the body. What Kant calls “desires and inclinations” are subject to natural laws, not to the laws of freedom. What Kant calls the will — the agency of human freedom, choice, and action — is the causal principle that is essential to persons. The function of will is to select, control, and direct emotions, passions, and desires rather than following their causal factors. On this point Kant seems to agree fully with the character Socrates in the Gorgias. The character Callicles proposes that the best human life consists of experiencing desires and passions and acquiring the means to satisfy them. Socrates thinks that is closer to a vision of hell. [Gorgias 492-493]
§ Their second strategy is to insist that computers require human beings to program them. The stock response to this claim by the proponents of computer equality is to claim that computers will take over the task of programming themselves. Kant would say that the most the computer could ever do is replicate the kind of laws that prevail throughout the natural world, and those laws must always be heteronomous — dictated by forces external to the computer, and it would not matter whether they come from the laws of physics, chemistry, biology, etc. or from a human designer. But human beings participate in another clausal order, the realm where will is the primary causal force. “As a rational being I count myself as belonging to the world of the mind. Only by regarding myself as an effective cause in that world do I call my causality a will” (p. 67).
· The moral power of Jonathan Swift’s satire concerning how we should treat poor people in human society comes from the same source as the moral argument concerning why human beings should be preserved and preferred to computers (or any other thing). Kant’s demand that we separate persons from things lies in the different kind of value each realm possesses. Things have what Kant calls a “market value.” This kind of value is determined by human needs and inclinations. A computer has a market price, which is determined by the laws of economics. The law of supply and demand is the best known aw of that sort. But persons do not have a market price. They have a different kind of value, what Kant calls dignity. We should not eat babies because they are persons, and as persons they have dignity. Kant’s categorical imperative is crystal clear on this point: “All rational brings are subject to the law by which individuals should treat themselves and all others never as a mere means, but always at the same time as ends in themselves” (p. 48). Persons have intrinsic value; things have extrinsic value.
· So, for Kant the difference between computers and human beings and computers is the essential difference between things that are only determined by the causal laws of nature and persons who are part of two causal realms, the laws of nature and the laws of freedom. Because human beings have both brains and minds, we can and should try to act according to the laws that govern minds — the autonomous laws of the mind that governs our will.
C. Scientific Materialism
· Modern science is often used as the basis for the value position that I will call “scientific materialism.” According to this view, our social, political, moral, and economic values follow from the way things have evolved naturally — according to the laws of biology.
· In the second half of the 19th century, the philosopher and sociologist Herbert Spencer embraced the idea of the “survival of the fittest” and used it to formulate a theory he called “social Darwinism.”
· Although few people have strictly followed Spencer’s extreme application of biology to society, during the first quarter of the 20th century a variety of movements developed that tried to build moral, social, political, and economic structures on a firm scientific (biological) foundation. The eugenics movement is one product of this way of thinking. It attracted a surprising group of people who assumed the burden of rethinking and redirecting human evolution. Karl Pearson, Theodore Roosevelt, H. G. Wells, George Bernard Shaw, and Winston Churchill are some of the names Matt Ridley links to the eugenics movement (See Ridley, Genome, pp. 287-294).
· Traditional forms of religion and morality were replaced by a new vision of heroic and omnipotent human beings who would take control of all aspects of life and lead the way into what Aldous Huxley called a Brave New World.
· In my next podcast (Human Nature # 5), I will consider some of the aspects of this way of thinking about values that captivated the imagination of so many influential people in the 20th century.
Podcast # 4: Artificial Intelligence
Albert A. Anderson
Copyright 2006
I. The Value Question
A. Behaviorism, Determinism, and Materialism
When I wrote “Minds and machines: A Modest Proposal,” I had three targets in mind: (1) behaviorism, (2) determinism and (3) materialism. I was especially concerned about the implication of those ideas for moral and political values.
A. Behaviorism
Searle’s Chinese Room is partly designed to question behaviorism — the view that there is no more to mind than its effects, its behavior. John Watson and B. F. Skinner embraced behaviorism and set the stage for many versions of that approach in the 20th century. Searle’s central point is that meaning is essential for human thinking, and meaning is a function of first-person language and immediate experience. The central difference between syntax and semantics is that syntactical structures can be instilled from the outside and observed from the outside. But semantics cannot be acquired from the outside. It is the essential element not only in thinking but also in acting and choosing. Philosophers as different as Socrates, Kant, Marx, William James, Gandhi, and Sartre, all agree on one central point: to be human is to choose and to act — to shape our own destiny. For that to be possible, we must be free agents. This leads to the problem of determinism.
B. Determinism.
Darwin’s theory of evolution, at least the way it is often presented, seems to undermine and threaten human freedom.
· Perhaps the best way to explain determinism in this context is by appeal to the 18th century French philosopher Pierre Laplace. He was steeped in Newtonian physics and was convinced that Newton’s laws of motion provided a full explanation of natural phenomena. This led him to claim that an ideal observer who knew the position of every particle of matter and knew the laws governing all their movements, that observer could predict as well as reconstruct all events in the universe. In other words, the laws of physics determine everything. Genuine freedom (free choice) has no place in such a universe. If the laws of chemistry are built upon the laws of physics, chemical laws are is equally determined. If the laws of biology are built upon chemistry, they are equally determined. If the laws of psychology are built on biology, they are equally determined. And so on throughout all of the natural and social sciences. No matter what we do, no matter how meaningless this might make what we falsely think are our decisions, there is no place for free will in such a system. Replacing human beings by machines (or any other products of evolution that are more fit to survive than we are) would be inevitable.
· But this leaves us in a classical dilemma: Either (a) we abandon freedom and the moral choices that depend on free choice, or (b) we abandon the causal laws of physics, chemistry, and biology and open the way for any unscientific or anti-scientific belief that can be imagined. Either way we seem to lose something a reasonable person cannot give up. Can we resolve this dilemma?
· This view of nature stands in direct opposition to the position taken by a philosopher who started out in physics and who was an older contemporary of Pierre LaPlace: Immanuel Kant. Kant was well aware of the dilemma posed by scientific determinism. For that reason he insisted on separating two fundamentally different realms to which we have access:
“There are two possible points of view from which we can see ourselves and recognize the laws governing the use of our powers and thus all of our actions. First, insofar as we belong to the world of sensation, we see ourselves subject to the laws of nature (of heteronomy). Second, we see ourselves belonging to the intelligible world, subject to laws that are independent of nature, not empirical but grounded in reason alone. As rational beings, and consequently as belonging to the intelligible world, we can never think of the causality governing our own will except as pertaining to the idea of freedom” (Immanuel Kant, Foundations of Ethics, Agora Publications, 1995, p. 66 [452]).
· Kant refused to give up either the scientific account of the empirical world of material things or the world of the mind that is subject to a different set of laws — the laws of freedom. His distinction between (a) heteronomy and (b) autonomy lies at the center of his attempt to avoid the horns of the dilemma that emerges from the modern scientific worldview.
o Heteronomy is derived form two Greek words: hetero (other) and nomos (law). The world of nature outside of human beings is ruled by laws that are alien to us — they are imposed without concern for our will — we have no choice in the matter. The law of gravity or the speed of light from physics and natural selection or the survival of the fittest from biology are heteronomous.
o Autonomy comes from the Greek words auto (self) and nomos (law). According to Kant, causality in this realm governs our will — or decisions, choices, and actions. To this, he says, “is tied the universal principle of morality” (p. 66). All actions by what Kant calls “rational beings” are grounded in autonomy. A rational being is any being that is able to think of laws that guide free choice and choose to follow them. This is also how Kant defines what it means to be a person.
· Persons and Things
o Kant would have no difficulty in explaining the crucial difference between computers and human beings. Computers are things, totally directed and controlled by heteronomy. Human beings are persons, guided by autonomy.
o The contemporary struggle to explain why persons can and should control the world is quickly and easily resolved in Kantian terms by showing the superiority of any being that directs and controls itself and a thing that is totally determined by outside laws and forces.
o Those who wish to distinguish human beings from digital computers frequently employ two different strategies.
§ In the first strategy they point out that human beings have feelings, emotions, passions, and desires, and machines lack those phenomena. But for Kant, this strategy is doomed to fail, because it locates human uniqueness in a realm that belongs not to the autonomous realm of the mind but to the heteronomous world of the body. What Kant calls “desires and inclinations” are subject to natural laws, not to the laws of freedom. What Kant calls the will — the agency of human freedom, choice, and action — is the causal principle that is essential to persons. The function of will is to select, control, and direct emotions, passions, and desires rather than following their causal factors. On this point Kant seems to agree fully with the character Socrates in the Gorgias. The character Callicles proposes that the best human life consists of experiencing desires and passions and acquiring the means to satisfy them. Socrates thinks that is closer to a vision of hell. [Gorgias 492-493]
§ Their second strategy is to insist that computers require human beings to program them. The stock response to this claim by the proponents of computer equality is to claim that computers will take over the task of programming themselves. Kant would say that the most the computer could ever do is replicate the kind of laws that prevail throughout the natural world, and those laws must always be heteronomous — dictated by forces external to the computer, and it would not matter whether they come from the laws of physics, chemistry, biology, etc. or from a human designer. But human beings participate in another clausal order, the realm where will is the primary causal force. “As a rational being I count myself as belonging to the world of the mind. Only by regarding myself as an effective cause in that world do I call my causality a will” (p. 67).
· The moral power of Jonathan Swift’s satire concerning how we should treat poor people in human society comes from the same source as the moral argument concerning why human beings should be preserved and preferred to computers (or any other thing). Kant’s demand that we separate persons from things lies in the different kind of value each realm possesses. Things have what Kant calls a “market value.” This kind of value is determined by human needs and inclinations. A computer has a market price, which is determined by the laws of economics. The law of supply and demand is the best known aw of that sort. But persons do not have a market price. They have a different kind of value, what Kant calls dignity. We should not eat babies because they are persons, and as persons they have dignity. Kant’s categorical imperative is crystal clear on this point: “All rational brings are subject to the law by which individuals should treat themselves and all others never as a mere means, but always at the same time as ends in themselves” (p. 48). Persons have intrinsic value; things have extrinsic value.
· So, for Kant the difference between computers and human beings and computers is the essential difference between things that are only determined by the causal laws of nature and persons who are part of two causal realms, the laws of nature and the laws of freedom. Because human beings have both brains and minds, we can and should try to act according to the laws that govern minds — the autonomous laws of the mind that governs our will.
C. Scientific Materialism
· Modern science is often used as the basis for the value position that I will call “scientific materialism.” According to this view, our social, political, moral, and economic values follow from the way things have evolved naturally — according to the laws of biology.
· In the second half of the 19th century, the philosopher and sociologist Herbert Spencer embraced the idea of the “survival of the fittest” and used it to formulate a theory he called “social Darwinism.”
· Although few people have strictly followed Spencer’s extreme application of biology to society, during the first quarter of the 20th century a variety of movements developed that tried to build moral, social, political, and economic structures on a firm scientific (biological) foundation. The eugenics movement is one product of this way of thinking. It attracted a surprising group of people who assumed the burden of rethinking and redirecting human evolution. Karl Pearson, Theodore Roosevelt, H. G. Wells, George Bernard Shaw, and Winston Churchill are some of the names Matt Ridley links to the eugenics movement (See Ridley, Genome, pp. 287-294).
· Traditional forms of religion and morality were replaced by a new vision of heroic and omnipotent human beings who would take control of all aspects of life and lead the way into what Aldous Huxley called a Brave New World.
· In my next podcast (Human Nature # 5), I will consider some of the aspects of this way of thinking about values that captivated the imagination of so many influential people in the 20th century.