Is Man a Machine?

Vladyslav Nazarchuk

Since at least the second half of the twentieth century, the idea of artificial intelligence has increasingly captured the popular imagination, blurring the line between human and machine. For example, in the groundbreaking 1984 science fiction novel Neuromancer by William Gibson, which founded the still-popular cyberpunk genre, the title “character” is an artificial intelligence which is able to embed within itself the minds of the deceased. The dead can thus live on forever in simulated environments inside the “mind” of the machine, with their memories and identity intact. Neuromancer offers such a place for Case, the novel’s protagonist, in a small bunker on a virtual beach alongside Case’s deceased girlfriend, saying “I am the dead, and their land. … Stay. If your woman is a ghost, she doesn’t know it. Neither will you,”¹ but Case ultimately rejects Neuromancer’s offer. While today, not everyone believes that a human mind is convertible with a computer program, it nevertheless may seem that such a reality is steadily approaching, especially with the recent “AI boom” which aims to replace increasingly more human-performed tasks with artificial intelligence.

However, a careful investigation of the issue using ideas from Byzantine philosophy reveals that a human is, at a fundamental level, much more than a machine, and a machine could never become a human regardless of how advanced its technology may be. By “machine” here is meant simply a physical, artificially-constructed system—whether an analog device, a computer, or a robot. First, to gain a feel for the secular debate surrounding the issue, three articles by contemporary thinkers on artificial intelligence will be presented, along with their main arguments. Then, a parallel account of human nature will be constructed from the perspective of Byzantine philosophy, mainly relying on the works of St. John of Damascus and Nemesius of Emesa, and overturning some of the erroneous assumptions of the contemporary thinkers along the way. Hopefully by the end, such an account will illustrate both the unique constitution and the inimitable powers of human nature, and show how vain and irrational the hope of replacing humanity with artificial intelligence really is.

Contemporary Arguments

One contemporary viewpoint regarding the question of whether a human is a machine can be found in a 1990 Scientific American article by John R. Searle, then professor of philosophy at the University of California, Berkeley, entitled “Is the Brain’s Mind a Computer Program?” In this article, Searle focuses on the specific question: “Could a machine think just by virtue of implementing a computer program?”² An affirmative answer to this question, or the view that “thinking is merely the manipulation of formal symbols,”³ Searle calls “strong AI,” and seeks to refute it using his “Chinese room” argument. In this argument, a person, who is not a Chinese speaker, sits in a room with baskets of Chinese characters, as well as a rule book containing a precise algorithm for transforming Chinese questions (bunches of Chinese characters) into responses. The rule book does not explain what the characters mean, only how to manipulate them based on their shapes; nevertheless, the person can run the algorithm in the rule book to pick appropriate Chinese characters to respond with, based on the questions fed to him. Thus, while from the outside, the room seems to give answers indistinguishable from those of a Chinese speaker, the person running the algorithm has no actual understanding of Chinese, and in fact no way to learn what the Chinese characters mean. Formalizing this argument, Searle makes the distinction between syntax (a language of symbols) and semantics (the meanings behind those symbols), that is, “between formal elements, which have no intrinsic meaning or content, and those phenomena that have intrinsic content.”⁴ As the Chinese room argument—in which a person can manipulate Chinese characters without having any understanding of what they mean—shows, “syntax by itself is neither constitutive of nor sufficient for semantics.”⁵ Since thoughts in human minds have semantics or meaning, while computer programs are entirely syntactic or formal, the important conclusion is that “programs are neither constitutive of not sufficient for minds,”⁶ or that “strong AI” is false.

To the question of what, then, makes the mind so different from a computer program, Searle replies that while a program is an abstract logical entity independent of its actual physical implementation, the mind instead is caused by the brain, in which “one is struck by the extreme specificity of the anatomy and the physiology.”⁷ In fact, “every ‘mental event’ … is caused by specific neurons firing in specific neural architectures.”⁸ Furthermore, “the brain does not merely instantiate a formal pattern or program … but it also causes mental events by virtue of specific neurobiological processes,”⁹ which a computer program, being immaterial, cannot do. Thus, Searle situates the human mind entirely within the specific biology of the brain, and in fact sees strong AI’s notion of an immaterial, independent mind, in the form of an abstract computer program, as “one of the last gasps of this antiscientific tradition,” which refuses to “treat humans as just a part of the ordinary physical, biological order.”¹⁰ As a consequence, however, Searle claims that it may still be possible for machines to think, as long as the physicality of the artificial brain has “causal powers (at least) equivalent to those of brains”¹¹—but the machine would need to be more than just a computer program.

An opposite perspective can be found in the article “Will My PC Be Smarter Than I Am?” by Ray Kurzweil, an American computer scientist, entrepreneur, and futurist, which was published in Time magazine in 2000. In this article, Kurzweil makes bold predictions about how future technology will surpass the current intelligence of the human mind. Kurzweil imagines the mind as a computer program running on neurons inside the brain; consequently, with inevitable advances in computer technology, both the brain’s “hardware” and “software” would eventually be able to be replicated. Kurzweil writes how “one cubic inch of nanotube circuitry would be 1 million times more powerful than the human brain, at least in raw processing power,” and that “by the third decade of the 21^st century, we will be in a position to create highly detailed maps of the pertinent features of neurons, neural connections and synapses in the human brain … and to re-create these designs in suitably advanced neural computers.”¹² After this, a sort of bidirectional transhumanism will ensue, in which machines become more intelligent than humans by “combining the complex and rich skills of humans with the speed, accuracy and knowledge-sharing ability that machines excel in,” while nanobot technology inside the brain would be able to alter human minds, and immerse them in a virtual world that will be “as realistic and detailed as real reality.”¹³ Kurzweil’s ultimate vision is that “by the second half of the 21st century, there will be no clear distinction between the two,” that is, between human and machine, and that “we will have a myriad of other varieties of intimate connection between human thinking and the technology it has fostered.”¹⁴ Implicit in all of Kurzweil’s predictions is the assumption that the human mind is essentially equivalent to a program running on a computer, which is precisely the “strong AI” position refuted by Searle. However, in his short article Kurzweil does not engage with Searle’s criticism of strong AI through the distinction between syntax and semantics, or Searle’s observation that “simulation should not be confused with duplication” and that “the computational model of mental process is no more real than the computational model of any other natural phenomenon.”¹⁵

A more developed view supporting the idea that machines can become human can be found in the 1997 article “Consciousness in Human and Robot Minds” by Daniel Dennett, then professor of philosophy at Tufts University. In this article, Dennett begins by responding, from a strictly materialistic perspective, to some common objections held by those who criticize the possibility of a “conscious robot.” The first objection is that “robots are purely material things, and consciousness requires immaterial mind-stuff,”¹⁶ which Dennett terms “old-fashioned dualism,” and argues that it is unscientific to single out the brain from the rest of natural phenomena, which so far have been successfully explained by physical science. The second objection is that “robots are inorganic (by definition), and consciousness can exist only in an organic brain,”¹⁷ which Dennett argues either falls under now-discredited “vitalism,” or that it is really of no consequence if a robot includes organic materials or not. Third, that “only something natural, born not manufactured, could exhibit genuine consciousness,”¹⁸ which Dennett terms “origin essentialism,” and refutes by reasoning that if two things are the same atom-for-atom, then the first thing’s origin does not by itself give it any “intrinsic properties” or “mystic difference” that sets it apart from its copy. The last objection to the possibility of conscious robots is that “robots will always be much too simple to be conscious,” but this is more an issue of technological practicality than a true philosophical obstacle, and is ameliorated by the possibility of “substituting relatively simple bits for fiendishly complicated bits”¹⁹ in robots instead of perfectly replicating the biological complexity of humans. At the basis of all of Dennett’s arguments is a strictly materialistic viewpoint, which holds that nothing exists in the universe except atoms in different configurations.

Dennett then provides an extensive description of Cog, an ongoing project of the AI Lab at the Massachusetts Institute of Technology to build “the most humanoid robot yet attempted,”²⁰ capable of learning from interactions with its environment, protecting itself from harm, and having an internal “motivational structure” in the form of “wants and fears, likes and dislikes”²¹ imitating those of a human. For Dennett, Cog is not interesting merely because of the technological problems its creation entails, but also because Cog, much like Searle’s Chinese room, can be used as an illustration or “prosthetic aid to philosophical thought-experiments.”²² For example, Dennett mentions the problem of “symbol grounding”—that the data structures inside an AI program have no proper reference to things in the real world—as a criticism of “strong AI,” which is similar to Searle’s criticism that computer programs can only have syntax rather than semantics. However, Dennett argues that the data Cog receives comes directly from the real world through its cameras and sensors, so “anything in Cog that might be a candidate for symbolhood will automatically be ‘grounded’ in Cog’s real predicament, as surely as its counterpart in any child.”²³ To the criticism that “nothing could properly ‘matter’ to an artificial intelligence, and mattering … is crucial to consciousness,” Dennett replies that, although Cog’s baked-in motivational structures might be “crude, simplistic, one-dimensional,” they are enough to at least rival those of simpler organisms, and rather shift the burden of proof onto those who claim that Cog’s pleasure and pain “is not real pleasure or pain, but merely a simulacrum.”²⁴ In general, Cog illustrates Dennett’s belief that “genuine embodiment in a real world is crucial to consciousness,” because “unless you saddle yourself with all the problems of making a concrete agent take care of itself in the real world, you will tend to overlook, underestimate, or misconstrue the deepest problems of design.”²⁵

The Nature of Man

After surveying the above arguments of contemporary futurists and philosophers of the mind, it becomes evident that the question of whether a machine can become a human is not so much about what kind of technology the humanoid machine should have, but rather about human nature itself, and especially about the nature of the human mind. If the mind is viewed either as a biological computer (Kurzweil), a computer with some physical extension into the real world (Dennett), or a more complex system with “causal powers” arising from the specific biology of the brain (Searle), it follows that the human mind can be replicated as either a computer program, a computer program running inside of a robot, or some more complex physical system, respectively. However, all three thinkers alike hold a materialist view in which the human mind is merely an epiphenomenon of the biological operations of the brain, rather than having an immaterial, independent existence. From this view, it would logically follow that a machine can, in fact, become a human, as long as it satisfies some equivalent physical properties which humans also possess.

Byzantine philosophy, however, gives an entirely different description of the human mind and of human anthropology in general, rendering many assumptions of these contemporary thinkers inaccurate. To begin, according to Byzantine philosophy, every natural object consists of a material body and an immaterial form, a view properly known as “hylomorphism,” rather than the “dualism” criticized by Dennett. While matter provides the “raw materials” for the object, the form (also called essence, nature, or substance) is the principle of unity for the object, giving it its overall shape. As Nemesius of Emesa reasons, “bodies by their own nature are mutable, dispersable and throughout infinitely divisible, with nothing in them remaining unchangeable; so they need something to bind them together … Every body needs something to hold it together, and so to infinity, until we arrive at the incorporeal.”²⁶ The form also determines what an object is at a fundamental level, containing those defining characteristics which are “presentative of the nature possessing it [the characteristic] and most proper to this nature itself.”²⁷ For example, St. John of Damascus writes how the form of a human includes the characteristic of being rational: “the rational cannot not be in man, because that which is irrational is not man. When it is present, it constitutes the nature of man; when it is absent, it destroys it.”²⁸ Other characteristics which do not constitute the object’s essence, but rather serve to differentiate one individual from others (such as hair color among different people) are called accidents: thus, in short, “being is divided into substance and accident.”²⁹

Now, the human mind is part of the soul, and the soul is the form of a living body; thus while inanimate natural objects only have a form, plants, animals, and humans have souls. Like all forms, the soul is incorporeal, and cannot be confined to the brain alone, “for things intelligible are not impeded by bodies, but spread throughout the whole body, wander and move in and out, and so they cannot be constricted by bodily place.”³⁰ Further, Nemesius explicitly refutes the view that the soul is merely an epiphenomenon of the material human body. Against the claims that the soul is a certain “attunement” or “mixture” of elements in the body, Nemesius argues that “the soul both opposes the body and imposes the word of command, as being its ruler; but an attunement neither leads nor opposes,”³¹ and similarly that “the mixture does not oppose the bodily desires, but even works with them … while the soul opposes.”³² Further, one can make an argument from priority: since the soul is what shapes the matter of the body, the soul must be logically prior to the body, while “if it is an attunement it did not exist before, but came to be afterwards when the body had been attuned.”³³ Finally, the claim that the soul is an attunement implies that the soul is mortal along with the body, which contradicts the Christian belief that the soul lives on after bodily death (although being bodiless is not the soul’s natural state), since “death is the separation of soul from body.”³⁴ All these arguments apply equally to both the claim that the soul is a mixture or attunement, and to Searle’s claim that “brains cause minds”:³⁵ both claims are just different forms of the materialist view that the soul is merely a secondary reality deriving from the operation of the material body, which is primary.

Therefore, a human being has components that are both material and immaterial, being “on the boundary between intelligible and perceptual being.”³⁶ Further, Nemesius shows that the immaterial soul even has primacy over the material body, writing that “the soul is in any case agreed by all men to be superior to the body; for the body is moved as a tool by the soul.”³⁷ This anthropology of man from Byzantine philosophy, as laid out so far, is already enough to overturn many assumptions held by proponents of artificial intelligence, and shows how futile the hope of turning a machine into a human is. The human mind, rather than being material or an epiphenomenon of the material, is part of the soul, or the incorporeal form of a human, and humans simply do not have the ability to artificially construct immaterial forms of living things, or to endow matter with these forms at will. Any kind of machine assembled from nonliving matter will only possess the forms of the materials making it up: since it was already shown that a soul is prior to the body rather than being caused by the body, even if a machine uses organic materials and is otherwise a perfect replica of the human body, it will never have a human soul appear in it, and thus will never become human.

To expand on the above, Byzantine philosophy holds that when a human is conceived in the course of the operation of nature, the human form (which is a universal that has existed since the creation of Adam) is enhypostatized or assumed by the individual human person—that is, is condensed or constricted to become an individual human soul, while retaining its universal character and thus its natural unity with other human souls. As St. John defines it, “that nature is called enhypostaton which has been assumed by another hypostasis (person) and in this has its existence,” “just as the human species, or human nature, that is, is not considered in its own hypostasis but in Peter and Paul and the other human hypostases.”³⁸ The soul thus provides an a priori internal unity to the human body, guiding the development of the fetus and all the life of the adult human. A machine, on the other hand, is put together from parts, “from the outside in,” and any unity it may have is only a posteriori, external. Thus, a machine and a human are irreconcilably different in their constitution. At best, a machine can perhaps imitate some of the accidents of human bodies—dimension, color, physical properties—but the essential, substantial characteristics of a human —growth, motion from internal impulse, rationality—remain inaccessible without the form. In fact, Dennett’s criticism of “origin essentialism,” the view that only something natural can have consciousness, is only possible because Dennett ignores the existence of the immaterial soul; whereas it is precisely a living thing’s natural origin that endows it with the soul necessary for life, whether plant, animal, or human.

Powers of the Soul

While the previous section focused on what a human is, an alternative approach to the problem of whether a machine can be human is to consider what a human does. This is essentially the approach of both Searle, with his argument that a computer program can only manipulate symbols rather than have semantics, and Dennett, who hopes that Cog can “perform a lot of the feats that we have typically associated with consciousness”:³⁹ whether or not a machine could perform the same activities as a human gives evidence of whether a human is a type of machine or not. Byzantine philosophy, as well, relates being to activity, in that the power for doing a certain activity is inherent in the object’s form or nature. St. John writes that “the nature of each being is the principle of its motion and repose… this is nothing other than substance, because it is from its substance that it has such a potentiality … of motion and repose.”⁴⁰ Motion, in turn, “is the actualization of potency as such”:⁴¹ thus, the potential for certain kinds of motion is actually present within an object’s nature, and this potential is actualized when the object moves in its specific way. St. Athanasius emphasizes that it is God who created and sustains the different natures of things and their corresponding motions, writing that

“He moves all things … and orders them according to their several nature. For simultaneously at His single nod what is straight moves as straight, what is curved also, and what is intermediate, follows its own movement; what is warm receives warmth, what is dry dryness, and all things according to their several nature are quickened and organized by Him.”⁴²

In the context of human nature, Nemesius summarizes these concepts as follows: “action comes from power and power comes from substance and is in substance.”⁴³

In particular, the human soul has three general parts: “they also divide up the soul in another way into powers, or kinds or parts, into the vegetative, which is also called nutritive and affective, into the sensitive and the rational.”⁴⁴ Humans share the vegetative part of the soul with both plants and animals: it works unconsciously, and allows the body to take in nutrients, repair itself, and grow. On the other hand, since nonliving matter does not have this ability, a machine generally remains the same as it was constructed initially, and needs to be manually repaired if it suffers damage or gradual wear and tear. The second, “animal” part of the human soul obeys reason, and “is divided into two, the desirous and the spirited,” as well as including the five senses and “movement according to impulse,”⁴⁵ or the faculty by which animals and humans move themselves. One one hand, Dennett’s claim that a robot like Cog “will be equipped with some ‘innate’ but not at all arbitrary preferences, and hence provided of necessity with the concomitant capacity to be ‘bothered’ by the thwarting of those preferences, and ‘pleased’ by the furthering of the ends it was innately designed to seek”⁴⁶ may seem to suggest that a machine can imitate the desirous faculty of the human soul, and Cog appears to have the ability to move itself according to these preferences. However, the difference here is that Cog’s “preferences” were explicitly programmed into it by humans, and thus both its “desires” and the movements it makes in accordance with these “desires” are actually entirely arbitrary and externally imposed, rather than originating from the robot itself.

The third part of the soul is the rational faculty, by which “man is linked … to the incorporeal and intellectual natures,” (including to the angels and to God), “in reasoning and apprehending and judging each matter, pursuing the virtues and cherishing piety.”⁴⁷ The rational part includes both a practical faculty, which governs things like virtue, arts, and crafts, and the contemplative faculty, which is divided into dianoia (discursive thought, pure logic) and nous (direct intuition). Searle’s argument that a computer “mind,” on the other hand, only manipulates symbols without understanding their meaning like a human does is certainly applicable here, while Dennett’s claim that Cog’s data structures are “grounded” in reality because the data comes from physical sensors, is refuted by the fact that Cog does not actually “know” that this data comes from the real world at all: Cog would behave exactly the same if those data inputs were simply fed to it from another computer. Even more importantly, the soul’s contemplative faculty allows the human mind to deal with universal concepts—ideal geometry, mathematics, beauty, truth, God—while a computer ultimately only understands certain sequences of electrical 1s and 0s, all of which are merely particulars.

While the above is enough to show that the human soul possesses many powers which machines do not have, perhaps the most important power of the soul is that of free will, that “the man who acts and makes is himself the origin of his own works and is autonomous.”⁴⁸ Put another way, this is the ability to make rational, autonomous choices, which Nemesius defines as “deliberative appetition of things that are up to us.”⁴⁹ Since it was shown that machines do not have true appetition because their “desires” are externally imposed, and that machines cannot deliberate because they are irrational, it follows that machines cannot make choices. Further, a characteristic of free will is that the human actions which come from it are contingent, meaning they can happen in one way or another, or, “the contingent is what is able to be prevented or of which also the opposite is possible.”⁵⁰ On the other hand, since a machine is built from nonliving matter, its operation is entirely determined by the laws of physics, and a computer will always give the same output whenever it is given the same input. In other words, for a machine, “in the presence of identical causes … there is every necessity that the same things should happen, and it is not possible for them to happen at one time this way, at another differently.”⁵¹ This again proves that machines do not have free will.

Conclusion

In summary, the problem of whether a human is a type of machine, or whether a machine could become a human, can be approached using Byzantine philosophy from two perspectives: what a human is and what a human does. The first approach reveals that a human consists of both a material body and an immaterial soul, which is a universal that is both prior to the body, and unifies, shapes, and gives it life. Meanwhile, a machine only consists of nonliving materials along with their nonliving forms, and its wholeness is only posterior to its parts. The second approach concerns the manifold powers of the human soul—its vegetative, sensitive, and rational faculties, including free will—and shows that a machine, not having a living soul, cannot possess any of these powers. Both approaches lead to the conclusion that there is an irreconcilable gulf between humans and machines: a human is not merely a machine, and a machine cannot become human. Notice, however, that this conclusion hinges on the existence of the immaterial, which tends to be denied in the contemporary discourse surrounding artificial intelligence. Thus, while a more enthusiastic view of artificial intelligence, such as that held by Kurzweil or Dennett, may treat the human mind as essentially equivalent to a computer, or a computer running inside a robot, a more skeptical view, as that represented by Searle, may recognize the unique biological aspects of the brain which resist being replicated artificially. However, since both of these perspectives are materialist, both ultimately miss the mark, not being able to appreciate the full extent of what makes a human being so different from any machine.

A fitting epilogue to complement the introduction of this essay, and to give some literary illustrations of some of the points made above, would be to relate a few scenes from a 1997 short story entitled “The Poetry Cloud,” by Chinese science fiction writer Liu Cixin. In this story, a “god”—a member of a race of advanced alien beings who have “mastered unimaginable technologies, and exist in the form of pure energy”⁵²—is challenged by Yiyi, a human poet, to write Chinese classical poetry which would surpass that of the famous poet Li Bai (701–762). The god believes this will be easy because “technology can surpass all”;⁵³ he takes on the form of Li Bai himself and begins to live a natural human life in the countryside, but still falls short of his goal. One day, this “Li Bai” and Yiyi are watching a beautiful young girl playing in the water by the river bank. Yiyi says,

“Just imagine: cut her open with a shape knife, take out her internal organs, gouge out her eyeballs and brains, pick out each of her bones, … and finally spread out a large piece of white cloth and lay out everything on it in anatomical order. Would you still regard that as pretty?”⁵⁴

When Li Bai expresses disgust at this idea, Yiyi goes on to explain:

“Nature in the eyes of Li Bai is the girl you see now at the riverside. The same nature in a pair of technologically-oriented eyes is the bloody components lined up in an orderly fashion on the white cloth. In other words, technology is anti-poetic.”⁵⁵

With this vivid image, Yiyi is really making the distinction between the whole and the parts, showing that humans’ ability to produce beauty comes not from the collection of the material components of the human body, but from the internal unity of human nature—which cannot be replicated artificially. Byzantine philosophy would fully agree, since, as mentioned previously, art is part of the soul’s faculty of practical reason, and the soul is precisely that which contains human nature and is the immaterial principle of unity for the human body.

Li Bai then decides on a different, brute-force approach: to write every possible poem within the strict forms of Chinese classical poetry. The gods store these poems in a large cloud of quantum chips, which “looks like cumulonimbus clouds in the Earth’s atmosphere which transform into majestic shapes that capture one’s imagination … giving off a sublime silvery glow, like a never-ending hyperconscious dream.”⁵⁶ Li Bai, however, laments that he has still failed his objective: “I have indeed composed the most supreme pieces of poetry by means of our great technology, but I have been unable to locate them in the Poetry Cloud,” since “technology met again with that unsurpassable obstacle in art, and a program which can appreciate ancient poetry is yet to be written.”⁵⁷ This illustration suggests that, as if by a mathematical law, merely being able to produce art counts for nothing if a machine cannot then distinguish between good art and bad art thus produced—a problem very relevant for modern-day artificial intelligence. Big-tooth, an intelligent dinosaur, concludes with what might as well have been the thesis of this essay: “Is the essence and nature of intelligent life really unreachable by technology?”⁵⁸