Will Man Become Obsolete?

I spent an hour on Fifth Avenue last week, just a visitor from Boston now, but awed, once again and as always, by the size and vitality of my native city. In the shadow of St. Patrick’s, I stood transfixed before the window displays of commercial spinoffs from computer technology—watches that play baseball and beep “Dixie,” radios thinner than my bankbook, $10 calculators representing a thousandfold advance upon the $400 device I bought with such a sense of modernity just ten years ago. It took this scale of densely packed, beeping, flashing, almost living and pulsating objects to force my reluctant paleontologist’s soul to a recognition that the revolution is already upon us—the most profound change in human life since everything from trains to television brought us all together. One block west, at Rockefeller Center, an inscription proclaims: “Wisdom and knowledge shall be the stability of thy times.” I wonder. Wisdom, perhaps, but….

Robert Jastrow’s The Enchanted Loom treats this revolution provocatively and with eloquence. He argues that computers will soon be sufficiently miniaturized and refined to become a genuine extension and improvement upon true human intelligence. An old chestnut proclaims that machines can never equal or even be, in any meaningful sense, at all like the human brain because improvements in computers only add circuitry, while organic intelligence is an ineffable, qualitative something that cannot, in principle, be matched by mere quantitative addition. This may be so, but I join Jastrow in deeming it not inconceivable that what we call wit, wisdom, brilliance, and insight need not have, as its material substrate, any more than a vast increase in the number and connectivity of circuits.

Hegelians and Marxists have long advocated the “transformation of quantity into quality” as a basic dialectical law about the nature of change. Graded inputs need not simply yield graded outputs. Instead, systems often resist change and absorb stresses to a breaking point, beyond which an additional small input may trigger a profound change of state. Water at 50 degrees Centigrade is not half boiling. A computer twice as big as another may not simply keep accounts twice as fast. Our metaphor about straws and camels’ backs reflects an implicit understanding that not all change is continuous.

The previous impediment, Jastrow argues, was not a technological inability to mimic the brain’s operation, but a limitation of size. If organic brains reach human capacities primarily by increasing the number and connectivity of neurons, then computers with enough parts may match our cognitive abilities. But the old vacuum tubes of first-generation computers would have required a behemoth several times larger than New York even to match an australopithecine. Miniaturization is the key to revolution. With ever smaller and more compact silicon chips, computers, Jastrow claims, will soon reach human capacity at human sizes. What then, he asks in a final reverie, would prevent a mortal human from emptying the accumulated evolutionary and social experience of his mind into a machine and achieving electronic immortality? Might silicon-based intelligence, albeit with an organic, carbon-based helping hand at the outset, represent “the mature form of intelligent life in the Universe?”

I do not accept all of Jastrow’s pronouncements on the merging of mind and machine, but I found this part of his book provocative and conceivably correct in outline. Unfortunately, it represents a small section at the end of a short book—an essay shorter than thirty pages, and not unsaid (if less eloquently) by others. The rest of the book attempts to portray this revolutionary development as the almost inevitable outcome of a continuous sequence of slow evolutionary improvements stretching back to the origin of life. As an essay in paleontology and evolutionary history, this discussion fails badly, and ends by portraying Jastrow’s hopes and theology, rather than the constraints of the world as biologists understand them.

Jastrow has tried to grant generality to his theme of computers’ transcending human intelligence by depicting the entire history of life as an inexorable and progressive march to increasing braininess; the carbon-to-silicon transition then simply completes a universal directionality. We have not witnessed such a reincarnation of the old chain of being since Teilhard’s Point Omega, and perhaps since Pope’s Essay on Man:

Mark how it mounts to man’s im- perial race
From the green myriads in the peopled grass.

Ever since the basic outlines of the fossil record were established more than a century ago, we have known how poorly the old chain of being matches the history of life. Its persistence as a metaphor and even, in Jastrow’s case, as an imposed “reality” merely reflects our unwillingness to abandon comfort in the face of evidence.

I criticize two aspects of Jastrow’s basic argument. First, even if life evolved as he states, this supposed directionality offers no guarantee of predictable continuity and advance in the man-machine transition. Yet Jastrow draws this message from the history of life: “It is reasonable to assume that human beings are not the last word in the evolution of intelligence on earth…. The history of life supports this conclusion, for it shows a seemingly inexorable trend toward greater intelligence in the higher animals…. If the past is any guide to the future, mankind is destined to have a still more intelligent successor.” But metaphor and analogy are not logical implication. Biological evolution is about ties of physical genealogy based on reproduction with error and natural selection. Computers do not breed. Any direction imparted to biology by its Darwinian mechanism does not translate to pathways of industrial change; a biological past is no sure guide to a technological future.

Second, and more important, where is the “inexorable trend toward greater intelligence” that dominates Jastrow’s biological vision? Most multicellular creatures are insects, doing very well thank you, and destined to outlive us, but not illustrating any temporal increase in intelligence to match their longstanding success. And each of our intestinal tracts contains more E. coli bacteria than the earth houses people. They will be with us at least until our intellectual essences enter those silicon chips. Life is a ramifying bush with millions of branches, not a ladder. Darwinism is a theory of local adaptation to changing environments, not a tale of inevitable progress. “After long reflection,” Darwin wrote, “I cannot avoid the conviction that no innate tendency to progressive development exists.”

Jastrow might argue that he is only considering the single pathway through the immense labyrinth of life’s bush that happened to lead to us. Even here I might reply that while we have a personal motive for special interest in (and affection for) this particular pathway, we have no right to regard it (or any other) as the essential direction of life. The pathways leading to aardvarks, anchovies, or artichokes are just as long, intricate, and biologically informative.

Even if we grant some validity to Jastrow’s special attraction for our pathway, where is the slow, smooth, and inexorable progress in intelligence that he sees? I don’t wish to sound perverse. I don’t deny that we started as single cells in the ocean, were once fish of limited intelligence, and can now build silicon chips, the agents of our future irrelevance. Still, we did not get from there to here on a predictable sequence of ever increasing mental power. Each lineage is a series of curious accidents, with long periods of stability (or numerous variants on basic designs), and occasional unpredictable changes that, in retrospect, we sometimes choose to call advances. What if primates had died in the Cretaceous extinction, or dinosaurs survived? Would horses now be praising good legs, or tyrannosaurs sharp teeth, as a universal criterion of advance leading inexorably to domination? What if glaciers had frozen the whole earth, and not just a part?

Jastrow seems to know that the overt appearance of our evolutionary lineage does not support his story of continuous and inexorable advance, so he tries to dig deeper by dubious inference. Most embarrassing is the observation that five-sixths of life’s history belongs to bacteria and blue-green algae that look pretty much the same at the end of these three billion years as at the beginning. Jastrow, undaunted, talks about “some three billion years of invisible progress.” In other cases, he distorts a story by ignoring diversity and only considering an abstracted archetype as the essence of a group. On supposed progress among the higher primates, he writes:

The monkey did not change very much from the time of his appearance, 30 million years ago, to the present day. His story was complete. But the evolution of the ape continued. He grew large and heavy, and descended from the trees….

What is the monkey? Monkeys come in more than one hundred species, from tiny marmosets in South America to baboons roaming the ground, to howler monkeys screaming and swinging through the trees. They are the most widespread and successful of primates. Apes are a dying afterthought (five species or so), with one peculiarly capable descendant. Moreover, with appropriate scaling to body weight, the small talapoin monkey has a larger brain than any ape.

Differences in interpretation aside, Jastrow’s book is studded with an unacceptable density of factual errors: for example, the first fishes did not have “nicely articulated bones in fin,” but an unossified pectoral flap: no reliable evidence supports the common claim that Byron and Cromwell had five-pound brains; the first australopithecines did not have brains twice as large as an ape’s when corrected for body size. Moreover, the book is simply too short to permit an adequate discussion of complexities in the immense subject it tackles—and it therefore often verges upon “popularization” in the unacceptable sense of “simplification to the point of distortion.” For example, in discussing dinosaurs, Jastrow writes:

In any given line of evolution, animals tend to increase in size from one generation to the next, other things being equal, because size provides security from attack, and thereby enhances the individual’s prospects for survivial. In each generation, the largest animals are the ones most likely to survive and produce offspring; their progeny, by the rules of inheritance, also tend to be larger than average; and so the tendency to increased size continues, until the limits of the available food supply are reached.

Here, Jastrow discusses a puzzling phenomenon known to evolutionists as “Cope’s rule.” Increase in body size is a common tendency in evolution, but it has none of the inevitability that Jastrow implies. The most careful study I know estimates that one-sixth of lineages (for Mesozoic clams) increased in size while five-sixths remained stable and none decreased. Moreover, Cope’s rule has no accepted explanation, though Jastrow flatly provides one. The best recent work on this subject argues that the frequency and extent of observed size increase is consistent with random changes, provided that initial ancestors are small and that more biological “space” therefore exists at larger sizes than at still smaller.

Why is Jastrow’s reading of life’s history so different from that advanced by most people who study the fossil record for a living? Do I detect a theological bottom line? Jastrow and a few other astronomers have tried to find God in the universe by reading the big bang as the cosmological equivalent of Genesis. I confess that I have found it hard to take this argument seriously. The big bang may have created our universe. In addition, by disaggregating the products of any possible antecedent universe into basic particles, it may have obliterated the history of previous worlds. (This paradox is no different from Hutton’s eighteenth-century contention that the cyclical nature of geological history wipes out the records of previous continents and oceans.) But an inability to reconstruct previous universes does not argue for their necessary nonexistence. All we can say is that we do not know; the issue of whether the universe contains enough matter to contract again (pulsating versus unique bigbang theories) remains unresolved. If scientists should not play God, they should stop trying to find God as well. The inquiry may be legitimate, but it is not part of science.

Jastrow, nonetheless, persists in his quest,¹ and now wonders whether the supposedly inexorable progress of intelligence does not point to a directing Intelligence. Is Paley’s Watchmaker about to raise his head after a century of well-deserved slumber?

When you study the history of life, and step back to look at this long history with the perspective of several hundred million years, you see a flow and a direction in it—from the simple to the complex, from lower forms to higher, and always towards greater intelligence—and you wonder: Can this history of events leading to man, with its clear direction, yet be undirected?

But the “clear direction” is an organizing principle in Jastrow’s mind. To his vision, paleontologists can only reply with Christ’s words to Pilate: “Thou sayest.”

If the history of life is not a tale of smooth progress, then our putative transition from carbon to silicon intelligence cannot be linked with paleontology to form a unified and grandiose vision of continuity and advance in the universe. I prefer to depict it as a true discontinuity in the earth’s history. (I also willingly confess my personal predilection for viewing sharp transitions and changes of state as an important part of nature’s panoply. What else does the output called “novelty” mean, however smooth the input?)

In one of his most famous aphorisms Freud argued for three great discontinuities in the history of Western science—each pushing humanity off a pinnacle of its cosmic arrogance: first, the revolution of Copernicus, placing us not in the center of the universe but on “a tiny speck in a world-system of a magnitude hardly conceivable”; second, Darwin’s revolution, which “robbed man of his peculiar privilege of having been specially created, and relegated him to a descent from the animal world”; third, in perhaps the least modest statement of intellectual history, Freud’s own revolution, “endeavoring to prove to the ‘ego’ of each one of us that he is not even master of his own house, but that he must remain content with the veriest scraps of information about what is going on unconsciously in his own mind.”

MIT historian Bruce Mazlish suggested in 1967 that computers would generate a “fourth discontinuity.”² Each previous event had left us one comfort. Copernicus showed that we live on a peripheral hunk of rock, but we could still believe that God put us there by fiat. Darwin proved that we had evolved naturally, but we still had our rational minds. Freud denied both our rationality and personal knowledge of our own minds, but we could still view our mental power as unique. As the fourth pinnacle crumbles, we must admit that a board of silicon chips might surpass all the cognitive power in our heads.

But Mazlish also points out that each discontinuity in our conception of ourselves establishes, as its cardinal substantive claim, an unperceived continuity in nature. First between the earth and other physical bodies, second between man and nature, third between mind and evolved matter, and now, finally, between man and the machines that he builds. Ironically, Jastrow wrote this book to establish what I regard as a false continuity in the actual, physical evolution of intelligence. Yet the idea that artificial intelligence might unify nature may well be sound in a more abstract sense. By forging a true discontinuity in the physical history of intelligence on earth, we may force ourselves to appreciate our own deep embeddedness in nature. Of course, any paleontologist knows that too deep an embedding can lead to oblivion. This, indeed is the paradox we may soon face.