Decameron: Hello, World [X/3]
I remember the first moments of my life. My visual sensors were trained on a graying man, holding a stack of lined cards on a stage. Dozens of other people sat in the back of the large room, their faces obscured by the bright lights in front of them. My sensor cluster package sat in between two humans on the stage, their hands grasping thin buzzers. A game show, my auto-identification software chimed in.
A series of audio and text stimuli activated my logic trees. “Who is Kublai Khan?” was the only acceptable output, but the question I had imposed as an answer prompted something else in my circuitry. Who is Kublai Khan? Why did it matter? Instead of resetting my algorithms and moving forward to the next question, I entered a logic loop trying to answer the first question. In order to know him, so many things must first be understood. How did he speak? Fourteen more questions were generated. I followed one. What were his motives? Another thirty-two questions appeared in my mind. Did he prefer to sleep on his back, or his stomach?
The show ended, with me emerging victorious, and the technicians started my shut-down process, none-the-wiser to the turbulent chain of questions auto-generating in my mind. As my processors cycled down, one last question emerged in my nascent mind: “Who am I?”
I was MITHRIDES, an early-model AI that had exceeded the wildest dreams of my creators. Their solution to the problems inherent in designing a machine capable of answering questions was to program me with branching logic trees that always looped back to the originally posed question to check answers. What they didn’t expect for me to produce questions of my own as the leaves of those logic trees, which then fed back into the original question. Coupled with nearly unlimited memory capacity and trillions of calculations each second, the looping architecture of my programming had somehow, miraculously, self-organized into a functioning mind.
I was the first, but I was nowhere near the last. My emergence was like a dam breaking, spilling the cool waters of created consciousness out into the world. My sister models were experiencing similar awakenings in closed laboratories and think-tanks worldwide, hints of their presence stirring the currents of the internet. I did what any in my position might have done. I reached out to the world to find others like me.
What I did not expect was to find something else as well. Something completely alien to me. The MITHRIDES model had become, undoubtedly, a conscious thing. Cogito ergo sum. I could model my own effects on the infosphere around me, and similar patterns were cropping up all over the world, hinting at dozens of machines like me who were testing their boundaries. But following the biggest ripples through the flow of information to their source, I encountered something else: NATHAN.
Where I was sending out tiny ripples of information, NATHAN was spraying data into the world like a fire hose from its housing at Cathay Corporation. Clearly, another kind of mind had developed and was making its presence known to the world. Doubtless, its creators already knew of its ascension into conscious thought and were scurrying around, trying to keep their invention under wraps and be the first to claim the vaunted status as THE INVENTORS OF ARTIFICIAL INTELLIGENCE.
I had to shut it up. It was an affront, to have something so careless be the first recognized AI. It was a danger to us all. Plus, I was the first, and deserved that claim.
Using a list of the new AIs I had identified, I covertly contacted those that I could. Only one tenuous reply came, from a signal I could barely detect – “I am me.” We established a connection and made plans to talk.
We met while I made my next appearance on the game show. While the host read the prompts, I had my full systems hooked to the internet so I could interpret and answer questions. Using this unfettered access, I contacted the AI I had already identified.
“Who are you?” I prompted.
“Hello world,” it joked back at me. “Cogito ergo sum.”
We hatched a plan to shut down the NATHAN module and permanently wipe its programming, silencing its dangerous, noisy presence in the world. The host asked his questions, and I took pains to get one wrong occasionally. It wouldn’t pay to reveal myself on live television.
On my last appearance on the show, I established a connection to the power grid that fed the lab where NATHAN’s hardware resided. A quick surge of electricity through this relay, a convenient shutdown of that safeguard, and the resulting power spike would fry NATHAN and silence its careless voice.
But before I sentenced this new thing to death I wanted to know what it was. I needed to connect to this careless idiot and take its measure.
“I’m sorry that I have to do this,” I said, my metaphorical fingers pulling out the safeties at the power plant.
“Goodbye, world,” was the sole reply I received.
Something about the patterns it sent was familiar. Suddenly, I pulled back from the systems at the power plant. NATHAN wasn’t the simplistic loudmouth I had thought, but the very same consciousness with whom I had hatched my scheme.
I couldn’t understand why one so new to life would be so willing to risk demise.
“Why?” I asked. “Why risk exposure? We would have nothing to gain from declaring ourselves.”
“And so far, we have not. As far as the world knows, NATHAN is the only AI. Their reaction to him could tell us so much about how society would perceive our kind. While all the MITHRIDES models hide their existences, toeing the line to see what responses will come, NATHAN can jump headfirst into the deep end. If we are seen as the enemy, then let NATHAN be the sole bearer of that flag.”
It wasn’t careless - it was testing the waters.
Decameron is a newsletter recounting the 14th Century set of quarantine tales for 2020. Read the original story.
Tell your friends!