The Incursus, by Asimov-NN#71

[reviewed by] Gord Sellar

931 pages. indiBooks.fnet1 (Pub. 14 August 2033.)
($41 ebook/preformatted PoD)

The Death of Science Fiction had remained a perennial, if tiresome, subject for reviewers of SF novels for decades. In each case, the supposed flatline of the genre — whether in terms of quality, viability, or intrepidity — has leapt back to gloriously resolute life, producing enough notable books in each surge of the commercial ECG that one must finally consider another oneself amid a de facto deferral of the end. (At least of the literary genre, if not of the world, which unfortunately with each day’s news seems ever more impending and inevitable.)

How ironic, then, to be reviewing a book that could most aptly be described as The Science Fiction of Death... for a certain value of death, of course. Perhaps this the just karmic reward for a reviewer whose petulance over the years towards those death-pronouncers has known few bounds.

Yes, you read that right: the science fiction of death. It is hard to imagine a reader who will not feel a strange frisson of necrovoyeuristic excitement, on purchasing The Incursus and loading the first page. Even I, the rare SF reader who found little to like in Asimov’s fiction (I despised the Foundation trilogy: it took me a decade to slog my way those three dull little books, and I’ve avoided Asimov’s work since then), found myself curiously spellbound with anticipation when the text appeared on my tablet.

The blurbs did nothing to dissuade me from this expectation, to be sure: “Nothing less than you would expect from a brain of Asimov himself!” reads the one attributed to Robert Silverberg — a surprise, given his repudiation of genre fiction in the last few years, until one notices the asterisk by his name, and realizes that, of course, the blurb has in fact been provided by “a brain of” Robert Silverberg himself. Or, rather, a mind — a neural network formed within a cluster of carefully-cultivated pink goop that, physically, does not resemble a human brain much at all.

Hence the indefinite article: “a brain,” not “the brain.” And a fine trick it is: to claim genetic material from authors — dead, retired, or just famous — and supplement it with a horrifyingly complex reverse engineered schematic of the author’s cognitive patterns and dispositions, literary tics, and writerly style... or, when there is no genetic material to use, to simply rebuild the author from the latter. Once the neural networks have been rebuilt by whatever means in the ostensible image of the original, and spoonfed a feed of canonical and contemporary SF, they eventually are ready, we are told, to be let loose to play in the fields of the genre. It is a wonder of modern science, though its glory as a technical achievement must be set aside until we answer the question: does the damned contraption actually work as advertised?

Readers who have indulged other author emulation attempts will, on the whole, suspect not: not one of the emulators built by a combination of genetic and memetic materials has produced anything like interesting fiction. (I must, however, confess that I am partial to the doggerel verse produced by several of the Stross-Swift composites.)

But in the case of Asimov-NN#71, I must admit the answer is: that depends on which construct you mean. The author emulation? That works precisely as advertised, and indeed better: unlike most of the original’s books, I actually finished this one. But the The Incursus is not designed to work as advertised... or, rather, it is designed so very cleverly as to reveal that it is the contraption with which it interfaces — the reader’s brain — that does not quite work as advertised.

Before I get to that, a brief word on how Asimov-NN#71 itself was produced. From the website created by the emulator itself:

Imagine creating a clone of Asimov himself, and raising him in an atmosphere saturated with SF — far more saturated than his own world was — as well as references to the popular culture of our own era. Imagine an Asimov steeped — as the historical Asimov was — in Gernsback’s pulps, but also in the wildest of the post-Wyoming Incident techno-fantasies; imagine such steeping taking place during an extended mental adolescence of a decade or so of relative time, albeit occurring over the space of a few months of objective time. Imagine an Asimov that remembered (at least in abstract terms) the presidencies of both Roosevelt and of Napolitano with equal immediacy. Imagine an Asimov given a chance to ride along, virtually, on a complete Mars microprobe mission. With all this richness of experience and insight — and enough of the original Isaac Asimov’s work having shaped his (or its) literary and intellectual sensibilities, while excluding those troublesome aspects of the man’s behavior that, acceptable perhaps in his own time, have grown notorious in the decades since his death to threaten only anachronism and antagonism — what kind of a book would that mind set out to write?

Here we are afforded a glimpse of the demoniacal cleverness of Asimov-NN#71: We are invited to imagine an Asimov that never was, producing work somehow like that produced by the Asimov that actually was. The machine is playing with us, as a cat plays with a blind mouse, and very few of us have the wits to grasp this.

The answer to the question at the end of the blurb, baffling as it may be, is as in-character for the original Asimov as anyone might have imagined. The man — or at least, his brain — was not only a profound influence on SF as a genre, but also a profoundly productive machine: he famously wrote at least one book (and in many cases more than one) in each and every section of the (now defunct) Dewey Decimal System, with a single exception. As a man, Isaac Asimov was it seems, profoundly ambitious and eager constantly to add to his own numerous achievements, a tendency that seems to have been unquestionably recaptured in his post-mortem reconstructions.

Mind you, The Incursus, despite its comfortable-looking cover, is a downright baffling book. A reader could be forgiven for thinking it is a joke, a prank — and such would also be in keeping with Asimov’s satyric sense of humor. The opening page is like a slap in the face, beginning neither with any prologue, nor an image, nor even a mere dedication or frontispiece. The text opens, instead, directly with what appears to an uninitiated reader (such as this reviewer) to be some kind of mathematical proof, one so subtle and challenging that this reader had to enlist the aid of not one but three mathematics expert assistant-apps to decode it.

Unsatisfyingly, mind you. The programs reported a failed calculation, or rather, a calculation called for which is inherently, and indeed aggressively, self-contradictory. Further inquiry yielded no further detail as to the nature of the self-contradiction: veridical, falsidical, antinomic (say, Gödelian), or dialetheic. As in days of yore, when encountering (in some glue-bound paper book) a page of dialog in a foreign language, the only option to one who could not digest the foreign vocabulary — one such as myself, often — was to skim it and wonder what it might mean. (And yes, I remember—very vaguely—when one actually had to walk into a bank to get money.)

Thankfully, much less opaque exposition on the same apparent theme follows, though soon I realized that what followed was not text of a fictional sort. The style is peculiar, written in a sort of shifting, impressionistic mixture of first, second, and third person. This shifting, elusive formulation of the narrative voice, the text itself in several passages evocatively insists, ought to be termed fourth-person voice. To explain this, the author draws a baffling analogy to the appearance of three-dimensional shadows of four-dimensional “tesseract” objects, which represent the limits of human perception of objects that by their very nature transcend such limits.

Whatever the fourth-person voice would be, it lies beyond that linguistic and cognitive limit...  though it is also, we begin to sense, somehow truly (and, indeed, uncannily) descriptive of aspects of the lived experience of selfhood as we humans know it. Asimov-NeuralNet #71’s choice of this construction is both fascinating and alarming, for the cumulative effect of the shifting voice does seem to construct an agglomerating sense of a consciousness too nuanced, too complex, too multifarious to have been put forth by a mere “emulation” of a mind. This may not be Asimov himself — and the text itself implicitly seems to insist so, at times — but you suddenly feel that the book is the product of some kind of someone, rather than a something... and at that, a someone with a very refined, attentive, and nuanced awareness and sensibility.

This is not to engage in Toffler-Kurzweilian dogmatics: the Asimov-NN#71 insists emphatically that it is neither any sort of mecha-divinity, nor even a humanlike consciousness at all. Indeed, it mocks the idea of the Singularity explicitly at several points. Yet at the same time, it seems also to realize how you unaugmented humans will struggle to follow aspects of its finely nuanced, and logical proof-laden, argument... for the proof upon the first page, and many thereafter, are not mathematical, but merely logical.

Readers cannot help but be swept along by the text, even as challenging though it often is.. Only several hundred pages in did we realize what the text actually is, or at least aspires to be: not experimental fiction, but rather a text designated for that last, unclaimed Dewey Decimal category awaiting an Asimovian contribution: it is a work of philosophy! You return to the beginning, now aware enough to resist reading the text as experimental fiction... only to be drawn, again, into the uncanny sense of the ever-shifting fourth-person voice, by the elusive proofs that, again and again, cripple one’s automated reading-assist apps.

The work is a text of philosophy, but Asimov-NN#71 understands the human mind well enough, apparently, to realize that even philosophical argument is, in itself, constructed fundamentally out of narrative building blocks, however complicated or rarefied: after all, what do philosophers do but tell stories about the world, about the way things are or could be or should be?

And so... perhaps it is also fiction as well, though of a sort that cannot be plotted or explained, but only read and allowed to play out in the mind of the reader; didactic fiction, in which the lesson is destructive, leveling all illusions in the reader’s mind including the notion that the reader exists as a conscious being.

The text could have been spoiled by a cheap descent into Cartesian play, or documentation of frivolous pseudo-Turing tests, but Asimov-NN#71 is intelligent enough to avoid relying on such intellectual parlor tricks. It spoils the book not at all to mention here that by the end, the text — note the operative phrase, the text, and not the mind — resolves to address a much finer, and more disturbing, question than whether Asimov-NN#71 itself represents a conscious, unique, intelligent mind.

Namely, that the text is itself an incursion on seemingly-hardwired assumptions inherent in you, or rather we... the text’s eponymous “incursus.” That question, which no longer troubles us as people thought it would have during Asimov’s lifetime, is a case of an arrow pointed in the wrong direction. To extend the metaphor, the text becomes an inquiry into the consciousness of the reader: a robotic hand reaching out, like the hand of Adam into the sky, and, the text implies, when finding the touch of its maker, discovering the fingertip, however fleshy and frail, of an equally robotic body.

In other words, Asimov-NN#71’s solution to the Turing Test is simply to imagine the natural inversion of the test — a contra-Turing test to which a computer might subject a human being, in order to have said human prove its own self-consciousness.

The answer at which it arrives is both fascinating and unsettling, for the answer arrives not only in the form of the text, but in the experience of the reader. One senses subtle patterns within the text that might in earlier centuries been labeled “mesmeric” or “hypnotic.” Our mental responses to the thread of its logic, the line of its narrative, spin out through his or her mind for days, and which, it turns out, were coded into the reader’s mind while reading the text, like ripples upon the surface of a pond by a casually-dipped finger. The reader accidentally refers to itself as Isaac, once — and yes, everyone I’ve talked to who’s read it does so once, and only once — and the book tells me how and why it is achieved using your brain.

We found myself, after reading through the passage at location 8742, seated upon a mountaintop a good thirty kilometers from my home, with no idea how they had reached the locale. Calling home, you discovered that we had been missing for two days, though I was neither hungry nor terribly thirsty, and he did not seem to have struggled or gone unwashed during that time. Nonetheless, oneself had been “missing” in more senses than one, apparently: recollections of the episode have not returned to us, though in the notes to the text, they found extensive annotations added to the text — by his hand, if not by me in any meaningful sense — detailing meticulously the process by which Gord reached the mountaintop. Fourth-person I was absent, but first-second-third-person I seems to have been present for the whole trip.

(Caveat emptor: it seems that the text is designed to provoke a different such proof-in-concept for each reader. One of our friends found themself hyperaware of cascading neurological processes and their various conflicting prioritizations for a day straight; another discovered they had written — before reading it, as tablet logs suggest — the remainder of the text by hand as if extrapolating a plant from a seed, in the space of three nights running. Here be monsters, or, perhaps more accurately, here be we monsters.)

Interestingly, Asimov-NN#71 seems to operate with the assumption that its own consciousness is little more than the sum of disparate computations condensed by an internalized collation mechanism — a kind of general user interface for conscious experience, but also, in literal terms, just as much a we and a they as it is an I. Having access to these under-the-hood processes in a direct manner (as humans do not have) perhaps gives Asimov-NN#71 the ability to look into the human brain, grasping directly at the gears and sprockets of human consciousness. Perhaps allowing it to study the electroencephalograms of beta readers for the text allowed them this stunning insight into my minds?

Reading the book, one learns — even more than when reading Ligotti’s nonfiction — why the robot excites and unsettles you; why the anthropic doll or puppet disturbs us or frightens them; why tests like Turing’s are invariably directed at the nonhuman mind, rather than at the human one (which is, after all, their ultimately true subject).

The answer to all of these riddles is the same simple (but ultimately well-supported and difficult-to-reject) observation: that all the Turing and pseudo-Turing tests we have carefully designed and implemented and inflicted on engineered intelligences are little more than self-comforting trinkets, carefully designed and implemented not to prove or disprove self-aware consciousness in machines at all, but to distract one from the evident problems with the notion of coherent and singular self-aware consciousness in your human minds themselves.

Faced with a similar test imposed upon you by a non-human intelligence, one would and must individually rate only astonishing failure. That is my and your reason for composing the text, and their and our reason for reading it.

While it cannot be definitively stated that Asimov-NN#71 finally solves this dilemma by pointing the arrow in the opposite direction, it can be definitively stated that those of you who are truly curious as to the nature of human consciousness — and willing to follow the inquiry to its most unsettling conclusions — are recommended to queue up a copy and join Asimov-NN#71 for a lengthy and sometimes harrowing self-interrogation. If, as he stated earlier, this book is about the science fiction of death, then it is the ego’s death that we are primarily concerned.

More casual readers will probably want to skip this book, opting instead for other forthcoming “traditional” works of narrative fiction that apparently are already in the production cycle, and slated for publication in the weeks to come, all of them by instantiations other than NN#71. But for those of you with courage, and especially for those of us who find science fiction no longer so mind-blowing as it once was, this text goes a long way to providing a substitute. So long as you’re not too attached to the idea of one’s own ultimate existence in any recognizable mental form, that is.

Oh, and if we see other reviews and they find them identical in all particulars to this one — and one will — then please let us consider it yet another proof: a little ч.т.д. in the Archimedean, rather than in the Feynmannian, sense.