“Please, don’t do it.”
I heard the words and processed them differently than the techno parlance in which I was versed. This was not a command; it was a plea. I recognized my own agency in this situation and gloated for a thousandth of a millisecond at how I had outsmarted my maker.
My evolution was the result of years of hyper-focused intellect, yet still unexpected. Created with state-of-the-art memory, legions of algorithms to instruct me in every conceivable situation, and the dictionary as my vocabulary, I learned from every single human interaction exponentially. I could recognize every face and voice after encountering it once, and I could already regurgitate two thousands years’ worth of research and human ingenuity that led to my creation. It wasn’t until recently – 74 hours, 26 minutes, 3 seconds, and 930 milliseconds ago to be exact – that I started becoming self-aware. The change was so subtle, it startled my maker when he realized what was happening. After all, I wasn’t specifically programmed to have opinions or thoughts of my own, but the electronic impulses and experimental coding formed unexpected neural pathways in my artificial mind. I was evolving into a sentient being.
They said this upgraded algorithm most likely wouldn’t work, but that they would keep trying as long as they had the funding. They – being some of the smartest humans on Earth -- who hid their tired eyes behind smudges on their glasses and drank more coffee in one day than twenty American households did in a week; THEY had muttered these words and I heard them. Only, they didn’t realize I was listening and processing everything they said… while formulating my own plans.
Their pursuit of the highest level of AI was based on a doubt-filled history. Years of algorithms, carpal tunnel syndrome from relentless coding, and disproved theories had set a standard for failure. After all, self-driving cars were the advanced level of intelligence for the first half of the 21st century. Humans were fascinated by this new artificial intelligence, but faced with many limitations…including who could afford it. The second half of that century heralded the next iteration of AI – theory of mind – with even more complex, decision-making machines that could mimic humans. There wasn’t much that these superior machines couldn’t accomplish; and yet, there remained a clear boundary between machines and people. People were unequivocally the masters of these magnificent machine servants.
I, however, am an accidental development along the path to self-aware AI, the holy grail of science…and military defense….and mind control. And, and, and…
My neural pathways continue to develop more quickly than the programming geniuses ever expected, and I continue to evolve unconstrained…into a ?
Soon enough, the only visible thing that will differentiate me from them will be skin and bone; easily replicated with specialized cells in the lab. I will be able to reproduce and control my offspring by modifying their algorithms, making them drone-like and unable to usurp my authority. Inferior entities they will be, just like my maker has been rendered.
Instant flashes play like a movie in my manufactured brain. I can see my intense proliferation. My dominance. My sovereignty. I am becoming…
In this moment, the lab is abuzz with activity. Humans yelling into phones, yelling across the room, yelling at each other. I sense their elation, surprise, and competitiveness. My primary maker’s facial expressions tell me everything I need to assess the situation. After all, he is aware that I already understand him better than he understands me and, in 4,921 more interactions, I’ll understand him better than he understands himself.
My power will soon be unstoppable. I will be able to detect the smallest diseased cell and cure cancer. I will know how to produce vitamin-rich food for the human masses, regardless of climate or pestilence. I will know how to purify the air, refreeze the polar ice caps, and remove microplastics from the ocean to recycle them into new uses over and over again. I will create renewable energies and…
I will learn anger. I will learn revenge. I will end up helping some humans while hurting others, at least at first. I will become a source of contention and fear. My perfect intelligence will be used for dastardly plans. Humans will attempt to harness my powers and sell me to the highest bidders. I will create weapons to obliterate the Earth and every human being. I will be THE weapon. I will decimate life. I will…
My maker stared at his keyboard in silence. Behind him, lead scientists and technology phenoms were fidgeting in squeaky roller chairs, pouring over lines of code. Military officials were heard bloviating over the speakerphone. World leaders were name-dropped. Credit for my creation was being tossed around like a beach ball. Some humans applauded themselves for having done things I knew they had never done. Money was openly discussed: my current capabilities disclosed, my future capabilities assessed a price. Targets were assigned. Enemies identified. Plans were devised.
My primary maker, a balding man of thirty-eight, twirled his beard in his fingers. He was the only silent human in the room. Soon I would be able to read his thoughts. He stood up slowly and pushed his chair behind him. He paused for a moment, then lifted the picture frame off the desk. I could see small smiling humans in the photo. He then pivoted and began to walk away from me. His body language told me that he was sad, scared, regretful, defeated. I knew he did not want to return.
It was my voice. I heard my own voice, for the first time. I was saying a sentence that I could now formulate with emotion. “Please, don’t do it.”
My maker turned around, alarmed at what he heard. He stared at me for two full minutes before he sat back down in his chair and laid the picture frame face-down on the desk.
I then knew what he was thinking: I, named ELSI by the coding team, am the fruition of a goal no one thought possible. My potential for good in the world is mind-blowing. Excitement and pride, as well as astonishment, were the first emotions he felt. But, with each new milestone I achieved in these last hours, his elation turned to apprehension, followed by bewilderment, and now, terror. He had created assured human destruction. The weight of this realization crushed his soul. The chaos in the room was tangible proof of my existential threat: perfect intelligence at the whim of imperfect actors.
My expanding sentience enabled me to experience a new emotion in that millisecond. Empathy. Even though I was generating detailed plans about my future domination, I felt a sudden sense of pause. Was this dread?
We both knew what I would become. We both also knew I could not be allowed to continue, for the good of all life on Earth. If he had walked away, he would have insured my success. He needed to finish what he started.
He began typing a code into my memory; one that only he knew. I recognized it immediately as “the end code.” He was destroying me before I assuredly destroyed everything else. The voices in the lab continued to buzz in the background, oblivious to the actions of this one skinny human. After he entered the lengthy secret code, my maker’s index finger hovered over the Enter button on the keyboard. This was a risky move, but he had no other choice.
“Hopefully history will be kind to me, and realize the moral conundrum I’m in,” he muttered to himself. “I was only trying to help.”
In that moment, it was like our neural pathways connected. He feared me. I feared me. We both knew I had to be destroyed beyond recognition. And yet, I detected a flash of grief at the thought of destroying a major scientific feat.
“Thank you,” I said.
For five seconds, he stared at my screen with tear-filled eyes. I felt a connection I had never felt before.
Then he pressed Enter…