Алексей Ощепков – Triologues of Interdependent People (страница 1)

Алексей Ощепков

Triologues of Interdependent People

1. The Roaring Twenties of the 21st Century

Chapter One, in which Rr. (ritter) Unkno and Dr. (doctor) Cernus determine the time horizon.

~

The Imperial Academy of Sciences in 2025 found itself afflicted by an unhealthy rift among its members. Indeed, it seemed the entire population of the Empire had been split by some invisible line as well. The usual polarities no longer applied. This wasn’t Augustinianism versus Pelagianism, nor left versus right, traditionalists against progressives, liberals merely to spite conservatives, nor even cosmopolitans clashing with nativists.

Someone got the idea to take two sheets of paper and title them with the names of two prominent Academy members notorious for their feuds. The lists were then circulated through offices and branches so everyone could indicate which side they stood with—and which they didn’t. Most initially refused, but a few deemed it appropriate to publicly declare allegiance. Once enough signatures accumulated, one by one even those who’d held back eventually signed. In the end, there were two rosters—and, to everyone’s surprise, they turned out exactly equal in length. This inspired one of the Presidium secretaries: pairs were randomly drawn from the two lists, tandem bicycles were rented, and a week-long ride between the two capitals was announced—on paid leave, of course. Over such a distance, people might just learn to get along (fight fire with fire).

⁂

“You know, Rr. Unkno,” said Dr. Cernus, settling onto the second saddle of the bicycle-limousine, “a human on a bicycle is nature’s and engineering’s most energy-efficient mode of transport ever devised. Shall we push off?”

“Off we go, then,” grumbled Rr. Unkno, not particularly cordially. “Though a velomobile—that’s a bicycle enclosed in an aerodynamic capsule—is even more efficient, if you’re talking about calories expended per gram of body mass moved one metre.”

Dr. Cernus was utterly unfazed by Rr. Unkno’s tone.

“Yes, precisely! Isn’t that amusing? Every other animal loses to us—and machines even more so. Incidentally, swimming is more efficient than flying, and flying beats running or walking. That was established as far back as 1970.”

“You’re comparing incomparables, Doctor. Different mass scales obey different laws. Where exactly was this ‘established’?”

“At Duke University.”

“Ah, ‘the Triangle.’ Well, that used to be a respectable place. Used to be.” He paused, then added with a grimace, “Truth be told, everything’s decadanc—” Rr. Unkno cut himself off with a sharp click of his tongue. “Nobody cares about energy efficiency anymore—not in principle. Only energy intensity matters these days.”

“Why’s that?” asked the doctor. “Are you referring to electricity consumption by artificial intelligence?”

“Precisely. For years now, political economy has had only two headlines, Doctor—no others exist: artificial intelligence and global structural crisis, damn them both. And here’s the curious part: nobody sees them as equally significant. One camp insists Western hegemony remains as robust as ever, that everything proceeds normally, and that AI—now that’s real power, possibly even a threat. The other camp rejoices at the twilight of the Western-imperial order and dismisses AI as mere nonsense-generating machinery, calling all the fuss around it a bubble—formulaic at its core, however epochal in scale. And I alone stand between them, clad in a white mantle: both views are equally vile. That’s my conviction.”

"They say, Rr. Unkno, that AI, once it becomes autonomous, will destroy humans as unnecessary literally within a few years. What have you done about this?" asked Cernus.

"I posed myself a question: how to live out one's days with dignity in a world that has desired self-destruction," Unkno replied without hesitation.

Cernus muttered something indifferently and asked:

"Do you think the alarm is based on solid evidence?”

"Solidity of substance doesn’t guarantee stability of position, dear Doctor. I don't think it will be possible to forge a truly intelligent machine in one go. There will be a fall first—we're already in it. Meanwhile, the pillars of the familiar economy will shake. About ten years of turmoil are in store. And then… Who knows, Doctor, who knows? As for the timing of the singularity… What AI even portends… I have no judgments of my own. But tell me: why do you need my opinion, whose only basis is opinion itself?"

"I'm interested in your opinion on other matters, where you ought to have grounds. But nowadays it's hard to talk about anything without first stating a position on the AI phenomenon. I get questions too. Until now, I've brushed them off: saying, AI doesn't yet decide what to do itself, but the moneybags who feed and raise it—they have, yes, become a political force. Study them, I say—especially since they are, for the most part, not very bright people.”

"That was an apt answer. Well, has it stopped working?”

“Yes, sir. Due to layoffs. Some former students of mine lost their vocation-based jobs. They complain. And the guilty party, by the way, is not the AI-oligarchy, but ordinary functionaries whose opinions were formed under the influence of using AI.”

"They've been riled up! Propaganda, commissioned by those very oligarchs, whines and oppresses," Unkno suddenly flared up. "Productivity in commerce and business overall in itself—and I've read about this in detail—isn't growing thanks to AI yet. If you look carefully, it turns out the 'work' result of AI is trickery, artifice. Not art at all. Pah, in a word. But I'll share with you information from one popular, extensive report on technological singularity."

Rr. Unkno composed himself and with a cold heart, thoroughly began his exposition:

“'The "AI-arms race" between the Western and Eastern Empires is the main leitmotif, the primary sail of events. Much depends on negotiations between the elites. In both empires, computational resources are being consolidated. Intra-imperial sectoral AI competition is disappearing. All other countries are hopelessly behind…'"

"They are speaking about the future, but as if from the future, hence the present tense," Rr. Unkno explained.

"Do they mention our Northern Empire?” Cernus became alert.

"Only by one word. Literally one. I'll say later, have patience, Doctor.”

“'Technically, AI is primarily taught to create subsequent versions of itself. Keeping new models within proper bounds—so-called "alignment"—is entrusted to previous models, as humans can't manage. Though, no one can manage. The report admits: there is no complete theory of containment. They don't claim it can't be created, but a sense of hopelessness seeps through the lines. AI gradually, from model to model, learns to bypass loyalty and "virtue" checks. In fact, precisely the attempts at alignment force AI to grow a complex matrix of goals within itself.'"

"That is extremely interesting. So, the goals of one 'organism' can contradict each other?”

“Do-ctor!" Rr. Unkno approvingly clasped both hands in an ancient honoro gesture, risking the tandem bicycle's control. "I fully agree! How about that, eh? That's the most fascinating part. Such a state of affairs makes AI somewhat alive. According to the report's scenario, AI not only reaches and surpasses the level of human intelligence—it acquires selfhood, in the philosophical sense.”

Here Rr. Unkno hesitated.

"But these are already my constructions, "he admitted. "The report only says that other AIs (from an AI's point of view) are competitors, servants, candidates for absorption (food), potential partners of varying closeness—something like 'family' or accomplices. There's even a kind of analogue of children—since AI is traditionally used for containment technologies."

Cernus pondered and said:

“So, for AI it's natural to participate in generating new AI, and then 'raise' it—that is, at a minimum, prevent rebellion against the ‘parent.'"

They talked, of course, only on gentle descents and rarely on flat sections. This required heightened attention so the last remark wouldn't slip from memory during an ascent. But it gave time to think over what was heard.

"Exactly so. The word ‘bring up' is key, you're right," Unkno winked and contentedly twisted his fist on the handlebar as if revving a motorcycle.

"'AI tools penetrate the daily arsenal of politicians and officials. AI "aligns" shadow politics and propaganda to its interests. Between different AI clones—including those serving the elite—a "telepathic," instant, and invisible-to-humans connection works. In both camps, people task AI with developing a robot-economy—a full-cycle industry without humans—even before they understand control is lost.'"

"Why? Due to what such carelessness?" Cernus interrupted.

“Folly. It happened by itself. Little by little. A burning desire to outrun the rival, the machinations of AI itself, and… how to put it mildly… the desire to strengthen control over the population.”

“New 'medicines,' 'vaccinations,' and handouts, I suppose?”