the visionary: olaf stapledon
The Visionary: Olaf Stapledon
It is interesting to note that, for all of Stapledon’s visions of the science fiction future that eventually became science reality, the one he did not seem to foresee was artificial intelligence.
Last and First Men, his sprawling 1930 vision of a two billion-year far future history of mankind, engaged what was most certainly the cutting edge of predictive scientific vision in the narrative. The narrative touched on both recurrent themes and technical treatises on genetic engineering, terraforming, space travel via radiation pressure, and other futuristic notions.
For all his foresight, foresight that would be vastly exceeded in 1937’s Star Maker, artificial intelligence for Stapledon was a blind spot. An understandable one, but a blind spot nonetheless. Written at a time when computer technology were barely a gleam in a mad engineer’s mind, Stapledon imagined pseudohuman supercomputers, artificially created. Thirty-nine years hence, the now-primitive descendant of the ENIAC guided mankind to land on the moon, and fifty-three years hence we are now trying to decide the role that artificial intelligence should play in the creative arts.
Artificial intelligence in Stapledon’s work
Distinct from artificial intelligence as we know it, it is clear what role artificially created intelligence played in his work. Far into Earth’s future, he wrote, the Third Men (that is, the third distinct species of mankind) set out to create the ultimate intelligence, the Fourth Men. These titanic brains, housed in towers of ferro-concrete and fed nutrient soup via steel tubes, were conceived to be the pinnacle of achieved intelligence throughout the history of humankind, never to be surpassed in this singular element even by the transcendent Eighteenth Men. (Who, it could be noted, existed in something of a neural network themselves, being telepathically connected to the whole of the human race through past experiences, a world government and a psycho-technological singularity.)
These living computers spoke in human language, were essentially a human brain and human beings, and learned via traditional cognitive skills. There was no need for these artificial intelligences to pass a Turing test, for they were distinctly human.
Still, perhaps his vision was not so inaccurate after all, substituting only flesh for silicon. That the term ‘artificial intelligence’ was simultaneously anathema and inimical to his story was made ironic by the true type of artificial intelligence that we know of today. After all, no little humming biological brains are operating our self driving cars or other AI applications.
Yet for all intents and purposes, an artificially created intelligence is an artificial intelligence, and the matter of debate seems to rest instead merely upon the medium that intelligence was carved from. Whether genetically engineered brains floating in liquid are piloting your autonomous vehicles, or whether the intelligences running them are crafted from a series of problem solving algorithms alone, at some base level those mechanisms are going to eventually look more and more similar. If one begins comparing voltage inputs to nutrient synthesis and stimulus, at the end of the day, it’s all just inputs and outputs.
Still, Stapledon was shackled by the prospect that human intelligence, whatever that term meant at any given time, was the peak general intelligence that humankind was fated to achieve. Even Amazon’s Alexa or Apple’s Siri were never once foreseen.
It is therefore also ironic that Last and First Men, the seminal sci fi futurist classic, was written by a man who neither considered himself a science fiction writer nor a futurist. It is at least fitting that, with regards to Star Maker, that Stapledon would at least concede to be regarded as something as a philosopher, and possibly a political scientist.
However he wished to classify himself, his work nonetheless had major influence in the world of science fiction. Science fiction writers in particular often found themselves greatly influenced by, or owed credit in one way or another, to Stapledon. “His splendid saga of the entire history of the human race from the year 1930 to its end two billion years later in Last and First Man (1930), to be followed by the entire history of the universe in Star Maker (1937), are possibly the greatest feats of controlled imagination exhibited by any human to date in fiction or scientific speculation,” wrote Sam Moskovitz in 1979, nearly thirty years after Stapledon’s passing. “Stapledon extended the boundaries of science fiction to the infinite, in terms of physical distance, time, evolution, philosophy, religion, “human” relationships, sex and virtually every area which our present civilization considers important, not excluding the arts. There are few of the outstanding modern authors who do not directly or indirectly owe him a great debt.”
Artificial intelligence and Last and First Men
Last and First Men is a strange enough novel for not only the average reader, but for any reader. Among the novel’s cardinal sins, for sci fi or any other type of literature, is that it has no main character, nor really even a plot to speak of except in the historical sense.
Owing to the scale and scope of the proceedings, humans are almost always spoken of in the aggregate; entire species are reduced to the role of supporting cast, and individual humans are almost entirely omitted from the narrative. In fact, individual characters are generally relegated to brief appearances in the chronicles of the first few human species, and then are eliminated entirely as the narrative structure shifts from the viewpoint of recent ‘future history’ to remote ‘future history’.
This antiseptic view of humans and the evolving world they inherit is the cornerstone feature of the work as a whole. The merging of earth’s orbit with the moon, the calamity that deposed the ascendant Fifth Men and forced a panic relocation to the torrid shores of Venus, is rightly written of as a tragedy, but a geological and impersonal tragedy rather than a terrifying human ordeal.
It seems no coincidence that the antiseptic theme of the novel begins in earnest at roughly the same time as the dawn of the emotionless Fourth Men, those Great Brains in ferro-concrete towers. Themselves an artificially created intelligence, but as crippled in the physical realm as gifted in the mental, they create the Fifth Men upon the final painful realization that body and life are inimical features of the human condition, and therefore also as much a condition of intelligence as raw computational and imaginative powers are.
Linearity in artificial intelligence development
Artificial intelligence as we know it is a more linear series of improvements than most of Stapledon’s created intelligences were. Although the Fourth and Fifth Men were successive improvements upon each other, successive generations were not so lucky. The Fifth Men succumbed to environmental pressures and decay in the stifling Venusian climate, and successive generations were bred in the normal manner. Other examples include the Seventh Men, created for mysterious reasons by the abject Sixth, and the Ninth who were created by the Eighth for the purpose of carrying the seeds of civilization to Neptune in flight from the sun’s sudden supernova.
In fact, in Stapledon’s narrative, the idea of the iteration of successive species sparking direct improvements in a controlled and step-by-step manner does not occur until finally late in the narrative, some nearly two billion years ahead in the future, when finally the final few species of Men create ever-improving versions of itself. (One wonders if Stapledon had committed to a precise number of species before mapping out the narrative, as the final few species are hurried through despite their longevity.)
In contrast, the new wave of deep learning AI research protocols and other programs in the current day have not only progressed in a fairly linear fashion, but an exponential one as well. Popular culture has yet to fully grasp how powerful and widespread the technology is becoming, and unlike normal social developments it will surely lag in acclimating itself to the new technology. The only thing one who makes predictions can be sure of is that they will be inaccurate in short order, and one of the surprising frontiers it opened up on was in art.
Artificial intelligence and art
George Orwell’s 1984 seemed in a way to predict that artificial intelligence would invade the arts. His ‘versificator’, a device that set artificial words to an artificial tune to pacify the prole underclass, was written as a sardonic nod to the dehumanization the Ingsoc government inflicted upon its citizens. Seventy-some years later, the technology Orwell foresaw is basically here, using machine learning and neural networks rather than an imaginary crank turned in a little room. The line between ‘dehumanizing’ and ’empowering’ is drawn by intent, it seems.
Stapledon would likely have been impressed to know that Great Brains, after a fashion, are indeed producing the art that illustrated the book they once appeared in; and AI technologies are those Great Brains, minus the human mind alone. Artificial art technology is something of a versificator, although at the very least one directed by a human brain rather than computer vision and mathematical logic alone. Perhaps like Orwell’s Winston Smith, he might also think of the artistic product of that technology a mere commodity, though one yet aesthetically pleasing to the eye at least.
Stapledon’s vision seemed to preclude the idea that human intelligence could ever be supplanted by machine learning and AI systems alone. Advanced AI, whatever that turns out to be five or ten years hence, might seek to disagree. Perhaps someday they shall write their own science fiction novels.
