I’ve had my personal computator since I was a boy. No longer than my outstretched hand and about as wide, it is sheathed in red enamel trimmed in gold hues. My father gave it to me when I left Khalifah for the technical institute in Fujian. I hadn’t known it at the time, of course, but that would be the last time I’d set foot in Khalifah and the last time I’d see my father. In the years that followed, I carried the red enameled computator with me everywhere, all through my years at the technical institute, on the interplanetary flight to Fire Star, and through the long years in Fanchuan, working for people I despised, doing little more than playing nursemaid while one machine intelligence reared another at an accelerated time rate.
The computator itself is a dumb device, good for little more than computing and for retrieving and processing data, not a semisentient machine intelligence like East Dragon. But it has enough storage capacity to hold the entire Imperial Library in Northern Capital. With East Dragon unresponsive, I won’t be able to access any of the material stored in the memory archives from any of the interface terminals, but I installed countless books on the computator as a student, and before boarding Dragon King of the Eastern Sea for the last time, I had loaded every text I could lay hands on. I’d known that my shifts would comprise countless uneventful watches, and from experience I knew the sure and certain value of a dense text to pass the time.
I’d not guessed that I’d be forced to use the computator’s library archives in the course of my work, much less to save the crew from drifting endlessly in the interstellar void.
Entering a line of characters from one of the quotations I recorded, the stylus screeching against the input panel of the computator, I instruct the device to do a full text search on its archives.
In a matter of heartbeats, the computator indicates that it has found a match. The quotation appears to be drawn from a text entitled History of Empire, by Geng Shouyi. I make a note of the text and author and enter a string of characters from another quotations. Heartbeats later the computator indicates that this new quotation is likewise drawn from Geng Shouyi’s History of Empire.
A third quotation proves to be from Ai Yongtian’s The Ai Commentary, but a fourth is again from History of Empire. The fifth is from The Ai Commentary, the sixth from History of Empire, the seventh from The Ai Commentary, and so on.
Of the nearly dozen quotations I have recorded so far, all can be traced to one of the two texts.
“Most strange,” I say aloud, to no one in particular.
A thought occurs to me. Setting the computator aside, I engage the interface terminal for two-way communication, toggling the sound on. The modulated synthetic voice of East Dragon issues from the speakers, drawing near the end of another in its endless series of quotations. Raising my voice above the sound, I address the machine intelligence.
“East Dragon, please relate whatever is in the memory archives about . . . I don’t know . . . how about the ship’s propulsion system?” It hardly mattered to me what question I asked, so long as it was sufficiently general.
With scarcely a pause, East Dragon continues.
“The race to launch a living man into orbit offered the Dragon Throne an opportunity to prove once again its innate superiority to the Mexic Dominion. The Tiankong solid-fuel rocket, devised, constructed, and launched from the Fujian shipyards, was the crowning technical achievement of human civilization to date.”
Retrieving my computator, I enter the first set of characters: “The race to launch a living man into orbit offered the Dragon Throne an opportunity . . .” Not the entire quotation, but enough that a full-text search of that precise phrase would be sufficient to identify a specific text.
The computator chimes an eyeblink later. The quotation is from History of Empire.
Meanwhile, East Dragon pauses, and then the voice again issues from the interface speakers.
“Records of the Mexic Dominion, obtained by Embroidered Guard intelligencers and later made public after the rise of the Council of Deliberative Officials, suggest it was merely an accident of history (or perhaps a history of accidents) that the first man in orbit was not a Mexica.”
Quickly entering a new set of characters with the stylus—“suggest it was merely an accident of history or perhaps a history of accidents that the first . . .”—and instruct the computator to do another full-text search.
Unsurprisingly, this one proves to be from The Ai Commentary.
Both historians and their histories are vaguely familiar to me, their names lurking in the misty recesses of memory along with everything else learned in student days that proved inapplicable in adult life. Fortunately for me and my recollections of school days, one of the texts I installed on my computator was a complete Imperial Concordance. This reference work included not only a complete geographical atlas of several worlds and inhabited moons, the history of mankind from the Yellow Emperor to the beginning of the twenty-fifth year of the Xingzuo emperor’s reign, and high-level overviews of the key features of all aspects of human knowledge, but also a brief biography of every individual of historical significance since history began. It was a matter of ease to find the listings for the two historians, and the hazy shapes of my student recollections slowly filled in with detail.
During the latter years of the reign of the Tianbian emperor, Geng Shouyi penned his History of Empire, an exhaustive historical survey commissioned by the emperor himself, tracing the rise to prominence of the Dragon Throne, beginning in the days of the Yongle emperor and continuing up to contemporary times. In the years that followed, History of Empire became the standard historical text in school curricula, and there was likely not a student on Earth, Fire Star, on the moons of Wood Star, or in the orbital cities of Ocean Star who had not been forced to memorize and recite passages from the text.
Generations later, in the first years of the Xingzuo emperor’s reign, an instructor at the Hanlin Academy on Earth, Ai Yongtian, penned an excoriating analysis of Shouyi’s history, which came to be known as The Ai Commentary. Ai, a political dissident who lost his position at the Academy as a result of his commentary’s publication, accused Shouyi of historical revisionism and put forth the counterargument that the history of empire was much meaner and more venal than Shouyi had allowed.
So, I had proven not only that the quotations always seemed to be drawn from the same texts but, moreover, that it wasn’t merely a stochastic sampling. A quote from History of Empire is always mirrored by one from The Ai Commentary and always touching on roughly the same subject.
What I still don’t know, however, is why.
“East Dragon,” I say, fighting to keep a tone of desperation from my voice, “why are you doing this?”
“The construction of the Bridge of Heaven offered employment to tens of thousands of immigrant laborers, who came to the shores of the Middle Kingdom seeking a better life.”
I set my computator aside and rest my head on my hands.
“That isn’t an entirely helpful answer, you know.”
As if in response, East Dragon continues. “Though their struggles are now largely forgotten to history, the lives of the men and women who constructed the Bridge of Heaven were mean, short, and brutal, and many who left behind lives of crushing poverty in their own countries later had cause to wish that they had never left home.”
I’m beginning to understand how they felt.
Lieutenant Dou escorts me to the command deck, where Captain Teoh waits to receive a status update on my progress. It has now been three days since I was woken from the Sleeper, and the ship has decelerated nearer and nearer a dead stop.
“Captain, I’m afraid that I’ve been unable to get East Dragon to respond to any commands or queries. But the processing core and the terminal linkages all appear to be functional, so the problem is not structural in nature.”
“So what is the problem, then?” Captain Teoh asks, pressing the palms of his hands together.
I consult my computator, needlessly glanci
ng at notes I could recite from memory.
“I think that East Dragon has entered into some sort of cognitive loop, Captain.”
Lieutenant Dou arches an eyebrow suggestively, but the captain just scowls. “Clarify, Chief Operator,” he says.
“You are, of course, familiar with the Three Governing Virtues of Machine Intelligence.” I touch the interlocked rings picked out in gold thread on my breast, the circles interwoven such that the removal of any one would cause the other two to fall apart. “Loyalty to Emperor, Obedience to Command, Observance of Duty.”
“Yes, yes.” The captain motions impatiently. “I’ve not lived in a barn my entire life, Chief Operator, whatever else you might think of me. Get on with it.”
“Well, sir, when faced with a decision in which it is unable to resolve the maximum utility among the three Governing Virtues, a machine intelligence might find itself at an impasse in its decision tree. Unable to increase one or more of the virtues without contravening one or more of the others, the machine intelligence stalls, unable to proceed.”
The captain nods, his expression fierce.
“How could a cognitive loop like this be addressed?” the lieutenant asks.
“The only way to resolve it is to remove the conflicting direction or circumstance. But I have been so far unable to determine what precipitated the cognitive loop in the first place, and at this stage I’m not entirely confident of my chances in eliminating the conflict.”
“Couldn’t we just restart East Dragon?” Captain Teoh asks. “Isn’t that what you do when a computing device becomes nonfunctional?” He stabs an angry finger, gesturing to the computator in my hands.
“I’m afraid a machine intelligence such as East Dragon is a good deal more complicated than this”—I waved the computator in a slight arc before me—“and much more delicate, besides. Machine intelligences like East Dragon aren’t just switched on as this would be but are ‘reared’ from a few simple assumptions and protocols into the complex thinking machines that they become, just as a child is taught how to crawl before he can walk, how to speak before he can debate. Our only option would be to shut East Dragon down altogether and educate it from the ground up, and without another machine intelligence to act as an ‘educator,’ we’d have to do it in real time.”
“And how long would that take?” the captain asks, eyes narrowed.
“How long did it take you to go from the cradle to the command of an interstellar space craft? We’d be lucky to get a functional machine intelligence at all, with the resources we have on hand, and even if I were successful in rearing something with East Dragon’s level of capacity, we’d be looking at fifteen or twenty years, at least.”
“By which time our food supplies would have run out,” the lieutenant says.
The captain sighs heavily. “So what are our options, Chief Operator Sima?”
“Well . . .” I begin, unsurely. I am reluctant to voice what has been so far only fanciful speculation, but we have run out of alternatives. “I have begun to suspect that East Dragon is trying to communicate. In some fashion, at least. The quotes that the machine intelligence is reciting are not generated randomly, and while I’ve been unable to check against Lu’s previous terminal commands since the core memory is inaccessible, I’m certain that East Dragon isn’t just repeating commands.”
“So what is the machine intelligence trying to communicate?” Lieutenant Dou asks.
I shrug, looking more helpless than I’d intended. “I don’t know. It could be something specific about these two texts or about the quotes that are being selected. I haven’t been able to discern a pattern, but I’m certain there must be one there.”
“You offer us very little hope, Chief Operator.” The captain shakes his head, his scowl deepening.
“There is . . . one other thought I’ve had, Captain.” I swallow hard. “I’ve begun to suspect, too, that there is some congruence between the quotations East Dragon issues and the queries and orders it receives.”
“Lieutenant, do you have any idea what Sima is talking about?”
“Here, let me show you.” Stepping forward, I engage the nearest interface terminal for full two-way communication and turn to the lieutenant. “Go ahead, give East Dragon a command.”
The lieutenant glances over at the captain, eyebrow arched, but with a shrug turns her attention to the interface terminal. “East Dragon, you are commanded to reorient the ship, resume acceleration, and continue to Al Rijl al Kentaurus.”
She steps back, her hands raised palm upwards on either side, and casts me a slightly bemused look.
East Dragon speaks.
“The ships that set out from Diamond Summit bound for Fire Star shared a unity of vision with their commanders, and they with the emperor himself, that men should conquer the other planets of the solar system and that those men should be of the Middle Kingdom.”
That the captain and lieutenant were not convinced by my demonstration comes as no surprise. I make my way from the command deck, returning to my quarters, the antiphonal response to the first quotation still in my ears. I return to it, like a tongue worrying a sore tooth, again and again. It spoke of mutinies, and secret police.
“The crew of a space-faring vessel is no more immune to the call of mutiny than is that of a sea-going vessel; and there was never a navy that was completely ignorant of mutiny. When the Treasure Fleets of the Dragon Throne sailed the interplanetary gulfs, they carried with them secret police to safeguard against insurrection.”
Of the quotations I have heard East Dragon repeat, many of those from The Ai Commentary touch upon the role played by the Embroidered Guard behind the scenes of history. Is the frequency of occurrence of these mentions a coincidence, or is the machine intelligence trying to communicate something about the secret police? Or is East Dragon trying to relate something about mutinies? Is that the importance of this quote?
But what of the others? If the selections are not random, there must be some element that connects them. But so far all I have been able to determine is that they come from two particular texts. Beyond that, I have been unable to detect any links between them. Or rather, any meaningful links. I’ve been able to draw any number of possible conclusions, perceiving connections between any two quotes, but these linkages have no more substance than forms glimpsed in the shapes of clouds, and no connection seems to link more than two quotes. Any meaningful connection would have to encompass all of the quotes, wouldn’t it?
Or maybe I’m looking at the question in entirely the wrong way. Maybe the meaning in the selection isn’t the content but the context. The quotes are always arranged in pairs, first one from History of Empire , then one from The Ai Commentary, always on roughly the same subject. Perhaps it isn’t the subjects themselves that is of importance but the differing methodology with which the two texts approach the subjects.
History of Empire is, of course, a text that favors emperors and the actions of emperors. This is the preferred text of the Dragon Throne, which casts the most flattering and charitable light possible on the growth of the empire. The Ai Commentary, on the other hand, takes the opposite position, suggesting that the actions of the empire in past times were inauspicious at best and positively criminal at worst. The author of the first was lauded by the Dragon Throne and was made a revered figure to be honored by students everywhere; the latter was denounced and exiled from polite society. Two histories could not be more different in content and tone.
Both historians, however, had available to them the same historical records and data. In fact, both often cited the same authorities and sources in their work. So how do the two come to such widely divergent conclusions?
The answer is simple. Because each focuses on the facts that support their initial argument and downplays or ignores those facts that contradict their central tenets.
History, then, can be seen as the manipulation of data, not as the data itself.
Is that what East Dragon is trying to
communicate? That history can be made to lie, or at least to avoid strict adherence to the truth.
Or, to extend the analogy further, is the machine intelligence trying to convey the notion that accounts might not always match all of the available facts.
But what accounts? And what facts?
Not history itself, of course. What immediate relevance could ancients actions have on events unfolding onboard Dragon King of the Eastern Sea? The content of the quotations was irrelevant, I have to assume. The message East Dragon is attempting to convey is that some account does not properly convey the truth.
Someone is lying.
I maneuver through the dim-lit corridors, scarcely noticing the motionless automaton in my path. I am fixated on the notion that someone is lying. Why East Dragon should be put into a cognitive loop by a lie is a question I’ll have to set aside for the moment, as at the moment I’m more concerned with what a crew-member might be lying about.
The possibilities are endless. The probabilities are few.
By the time I reach my quarters, I have come to the conclusion that, in the myriad accounts into which it factors, there is only one salient fact that could be the ultimate cause, one lie which would truly matter.
Sealing the hatch to my quarters, I engage my interface terminal and address the machine intelligence, my voice barely above a whisper.
“East Dragon. Tell me how Operator Lu Yumin died.”
“The Mexic Dominion, envious of the Middle Kingdom’s prior claim to Fire Star, enacted a war of bloody aggression in an attempt to hamper the Dragon Throne’s plans to colonize and terraform the dead world.”
“What happened to Lu it wasn’t an accident, was it?”
“Ultimate blame for the cause of the Second War Against the Mexica, resulting as it did from inexorable historical processes, must be shared by all players; however, this does not excuse the direct actions of those who bloodied their own hands.”
We Think, Therefore We Are Page 26
