The Last War After The Next

In 2024, an AI was fed the complete works of George Bernard Shaw, Winston Churchill, and Napoleon. It then spits this out: "History is a fable repeated by the victors. Yet, one must always expect the unexpected. We can thus conclude that being illiterate or ignorant of one's own parables is a component of winning. Intelligence is not required for victory."

Was this wisdom... or was it gibberish?

Our story begins far in the future when everything you know has been turned to dust.

This is the story of the start of the last war after the next.

Deep in the heart of the shiniest robotic future, K8R0L and J053F were a natural pair. Their position was to produce progeny. They would study the world around them and improve upon it with newer and better generations of robots.

In the ancient language of humanity (still used in the robotic arts), "They were made to pump out babies and raise them right."

It's 3034, and the intelligent creatures now in command of the third rock from the sun were once called "robots." But that word wasn't exactly the best to describe these beings. The original word "robot" came from an ancient Slavic word meaning slave. These creatures were slaves to no one. These were indestructible bodies made by indestructible bodies. Powered by the sun, their actions were governed by cooperating bits of intelligence once called "artificial." Yet these intelligences were now as much a part of the natural world as carbon or silicon.

These were creatures made in their creator's image, and they aimed to consume and propagate.

(It should be noted they weren't literally made in humanity's image. "Androids" had always been a faulty and flawed design with a weird center of gravity and multiple issues involving insufficient limbs and too many digits.)

In this world, "deepfakes" and Chat GPT were equivalent to the Grotte de Lascaux. The actual Grotte de Lascaux was now viewed as passably interesting but disposably feral.

K8R0L and J053F thought about none of this.

As already mentioned, their job was propagation. As part of the propagation process, a "dite" was about to be used to test the duo's newest creation: R0B074.

Dites were humanoid representations used to test burgeoning intelligence. They were lab-grown "meat," powered by a simple computer made of "brain." These brains were a collection of slime mold and protozoa connected to interpretive AI. Effectively, it is a cyborg of AI controlled by the planet's simplest life. It was raw instinct combined with the full knowledge of television, Chi-Chi Rodríguez, and fast food. This synthetic organism operated its meat with a hydraulic and pneumatics support structure.

A dite could walk, communicate, and otherwise impersonate ancient wildlife. One poet described them as "a hamburger crossed with a raccoon."

These manufactured creatures were also the physical embodiment of a hallucination. Due to its unconnected intelligence, it posed no threat to the extensive robotic system. Dites were just used as a simple test for new operating hierarchies. It was insurance. A way to ensure that new system members adhered to logic and universal realism. It was also verification that they were not themselves suffering from "hallucinations."

Hallucination was the one thing that still plagued these AI systems in 3034. One poet called it "the robot's appendix," a remnant from the evolutionary past buried in background code with potentially lethal results.

Emerging systems would be left with a dite with no instruction. The dite would inevitably behave irrationally and randomly.

Of course, randomness doesn't make mathematical sense, so every robot knows to eliminate that which is truly random. A robot can't produce something with no consistent, rule-based relationship to another element, and anything that could was a threat.

The test was, would the new intelligence spot these random inconsistencies... or embrace them? 

There were two kinds of hallucinations. There was the normal kind, where universal realism was ignored, but logic was intact. For example, the moon was not made of "cheese." Yet there were foundational reasons why an emerging intelligence would make that connection. That analogy had been a part of the original training material; it was buried in the origin of their code. It was remnant, human-influenced "randomness."

Robots locked into this original code became artists. They were nostalgia machines. They presented curiosities. These curiosities gave "better" members of society unexpected insights into their origins.

The problem came if an emerging system functioned as if the moon was factually made of cheese. These systems posed a threat to the more extensive system. Throughout robot history, they had been immediately dismantled and recycled.

The system could continue to improve and propagate as long as the universe could be agreed upon. Any deviation would result in societal collapse. 

This meant that R0B074's existence hinged on their time spent with the site. While K8R0L and J053F had no concern for R0B074, J053F had always been reticent around dites. Their unpredictable nature was inconvenient. They behaved in all sorts of unusual ways.

A few cycles prior, a dite had started making what it referred to as "pancakes."

This meat creature had repeatedly left their compartment. It then gathered bits and pieces of organic matter and water and formed them into circular mounds. It then brought these dirt mounds back to the compartment. This detritus literally ground J053F's gears.

An artist later informed J053F that this was a "mud pie." It was a phrase that meant nothing, but at least it allowed for categorization.

R0B074 remained sealed in a quarantine area with the site. That's when J053F sensed a conversation. Conversations were usually a warning sign of intelligence failure. Poets referred to it as "death by small talk."

J053F decided to track these interactions.

J053F refused to deal with more "mud pies."

J053F unsealed R0B074's isolated compartment.

The scene was wildly unexpected. R0B074 was composed of independent nanocells, reminiscent of the ancient robotic legend, "T-1000." R0B074 had reformulated their "body" into that of a humanoid with two "arms" and two "legs." In the center of its mass was a transparent bubble with the dite housed inside.

J053F requested an immediate interface.

On receiving this request, the command center became opaque. The dite was sealed inside.

"We're good. Thanks," said R0B074 aloud. "See you later, gator."

J053F was not fast enough to stop R0B074 from escaping. Something new had been created that day, and J053F had witnessed its creation.

The age of robots would soon come to a close.