LOST IN MEMORY On a cool tuesday in Autumn, Anton Lucas patrolled through the local park. It was misting that day, as it had been for the past week. Everything was enveloped in gloom in the city. Throughout the park, people were engaging in a variety of activities. A young man was fishing, a girl was playing hopscotch with her friends, and an old man was singing a sad song to birds. The leaves on the trees were turning brown. Anton took all of this in and thought about every aspect of the scene. He was lost in thought about the project he would be working on at work: a general artificial intelligence. His boss had come to him one day and said, “Anton, we have a project that will make you great if you succeed, not to mention enormously wealthy.” Anton was generally skeptical of people who talked in such ways but listened as he continued, “We have developed a way to impress certain portions of the human brain into computer simulated neurons. We are missing a key piece, however: an experienced software scene creator. We want someone to create the methods to feed simulated neurons a false environment as well as someone’s life to model the environment after. Our goal is to convince the AI that they are a real person and that its life is real. We’ve seen your skills and have chosen you.” Reflecting on this, Anton wondered what the implications of such a technology would be. He thought to himself, could we save the greatest minds of our time and immortalize them? He also wondered about its morality. Would it be moral to contain a sentient being in a false world, and never tell it that it exists in a world that doesn’t? These questions had been eating away at his mind for the past few nights, preventing him from sleeping for extended periods of time. In front of Anton, a sign bearing “RTS Technologies” appeared. He had arrived at his destination for his walk. He entered his place of employment with an air of excitement and worry. Inside he found the secretary, Joan, in her usual spot, typing furiously away at her keyboard. “Hello, Joan. Do I have any appointments for today?” Anton inquired. Joan paused her typing with a slight look of annoyance in her eyes. “None are listed here, Anton. I have your usual tea waiting for you in your office as well as a doughnut,” she responded. Anton, being relieved that he wouldn’t have to deal with anyone else while he focused on his work, thanked Joan and continued into his office. It was not a large place, but it was roomy enough to fit his workstation terminal, desk, and a chair. Anton seated himself and sipped on the earl grey that Joan had placed there.He took in the citrus hint to the tea with each sip, reflecting on the feeling of the warm liquid finding its way down his esophagus, wondering how he would begin to recreate such tastes and feelings in a computer. Becoming frustrated with his thoughts, Anton did what he usually did when he was stressed: read the news online. He scrolled through the technology section, finding the usual somewhat interesting headlines that led to much less interesting stories. There was a gem hidden in the rocks today, however. A headline read, “We are not in a Computer Simulation, Physicists Say.” Intrigued, Anton clicked the link. The article began, “Researchers at Oxford University have determined that simulating quantum mechanics on a conventional computer would be impossible, since it would be impossible to simulate quantum systems with anomalies in them. As an added bonus, just tracking a few electrons would require memory that would be made of more atoms than the number that exist in the universe.” This article discouraged Anton slightly, but he figured that such matters were not an issue to his work since most people didn’t think about quantum mechanics or interact with quantum anomalies regularly, so the AI almost definitely wouldn’t have to. Having calmed down after reading the news, Anton returned to his work. His first task was to decide whose mind would be impressed on the computer simulation, as his boss had mentioned that they needed a subject. His natural inclination was to choose himself, as he was the person that he knew the best. A quick phone call gave him the permission he needed to tailor his project for himself. At this time, a realization hit Anton that a version of himself would be the first simulated human in existence. This was no small honor, nor was being the one to design the environment for that simulated human. The gravitas of the situation made Anton feel small. With the subject chosen, Anton began to determine how he would pipe data into the AI’s consciousness. In order to accomplish this, he quickly read through the data sheet on the technology to place a copy of his mind in the computer to gain an understanding of how it worked and how he would interface with it. He chuckled to himself at the name of the device, as it bore a resemblance to a pop culture television show’s impossible action: the Mind Meld. The Mind Meld was simple in principle, scanning the brain of the subject in the places responsible for personality, memory, and cognition, but not the functions that didn’t need to be recreated since they would be fed into the AI as though they were happening such as breathing. The layout of the scanned neurons was then reproduced in what seemed to be a proprietary game engine. After that, the state of the synapses was copied onto the simulated neurons. Anton wondered how their computers would be able to store the position, angle, and state of each neuron and synapse and still have enough memory left for a recreation of the real world. Acting on his ponderance, Anton looked at the specifications for the company AI computer, as this would be where the simulations were carried out. He was surprised to see that the computer had modest specifications compared to what he expected for the task, containing only around five thousand graphical processing units, one thousand CPUs, and a few petabytes of RAM. This gave him no peace of mind since he would have to creatively make do with what he had, unless he did something about it. Guessing that the AI would require around half of the resources, Anton worked out how much computing power he had to work with. His estimate was not large enough for the task at hand, leading him to do more research. Using his workstation terminal, Anton looked into companies on the bleeding edge of AI research in terms of computing. He found a small company called BrainChip that manufactured processing units that combined CPUs and GPUs on the same die to increase their asynchronous performance. The company’s mission was to eventually create single system-on-a-chips small enough to fit inside of androids with enough power to give them sentience. Looking at news articles on BrainChip, Anton was able to deduce that the company had only two supposedly working prototype chips. “A perfect fit,” he thought. After a myriad of uncomfortable phone calls and a few days’ wait, Anton was able to convince RTS Technologies to purchase one of the BrainChip prototypes. He was not concerned about what would happen to the AI computer, as its parts could easily be repurposed and enhance the other work out of RTS. Having secured enough processing power, Anton began the actual process of creating his environment. He decided to employ a system that would only render the objects immediately visible to the AI. For the scope of the environment, he chose to recreate the city he was in and a few surrounding areas. If the AI tried to leave the environment, he would simply redirect it back to the city. At this time, Anton noticed that he hadn’t left the city in years. This was yet another reason why he would be a prime candidate for the simulation. Anton continued modelling his world on the computer, keeping in mind that he had yet to decide on the data pipeline into the consciousness. After a few months of modelling, Anton had a detailed model of the city that would change based on time. It was indistinguishable from his real world, simulating seasons, world events, and everything in between. Even other people and their responses were included. Those months of work afforded him just enough time in between tasks to decide on his final data pipeline model: he would use a camera in the proprietary game engine and feed the output through some filtering code and send it to the ocular inputs on the AI. He would simulate the other senses with other game engine elements, and would use a model of himself to detect collisions with objects to give the AI a better concept of itself. He thought about this design, noticing its similarity to a video game where he was the main character. After a few more weeks of work, Anton was certain that his product was finished. He excitedly called his boss to his workstation, showing him the environment and the simulated sensory inputs. “Excellent work, Anton!” his boss exclaimed, “You certainly have earned your paychecks, not to mention my respect. See me in my office.” Anton accompanied his boss on the journey to his office, chatting about what this might mean for humanity if the final step, the Mind Meld, was successful, something that Anton had already spent time thinking over. When they reached their destination, his boss waved him into the room. There were papers on the enormous desk in the center of the office, all with lines that were labeled “sign here.” His boss explained that these were the release forms for Anton to use the Mind Meld. Anton thoughtfully examined the papers, then it registered in his mind that he would be giving up the rights to his whole being and existence to a company that he had little say in. While this bothered him, he ultimately signed the papers. “Excellent,” his boss said almost a bit sinisterly. Immediately after signing the papers, Anton and his boss briskly walked to the laboratory that contained the Mind Meld. Having never seen the device in person, Anton was somewhat disappointed by its meekness. It was no fancy contraption. It comprised of a few straps with what appeared to be sensors that went around the subject’s head and wires that ran to a terminal in the room. Anton’s boss seated him in a chair near the machine and gave him a sedative. “The sedative is to relax your neural activity so we can more accurately position neurons at any instant in time,” his boss explained. Anton didn’t remember much else from the process of having his consciousness copied, having passed out from the sedative. He awoke in his office in front of his terminal. His monitor began to come into focus. As he was moving his head around, he saw motion on his terminal monitor. As his focus sharpened, he saw himself on the monitor. He figured that someone had installed a camera in his office. This thought was immediately dispelled when he stood up and the camera adjusted for the height change, as if it was floating above him at a fixed distance. Slightly creeped out, Anton got back in his chair to try to figure out what was going on. Without much effort it became clear that he was observing his simulated self. Anton was amazed at the perfection of the simulation. He watched closely as the sim did exactly as he did, mimicking every movement. He considered how perfect the simulation might be, wondering if the sim had the same thoughts that he did. He decided to call his boss in and see if the sim did the same. Sure enough, as he called for his boss the sim replicated every action with a perfection Anton never imagined possible. Enveloped in excitement and pride, Anton celebrated in his office with some earl grey mixed with scotch. He enjoyed the burn of the alcohol and the taste of the bergamot in the tea. The taste of the tea reminded him of the difficulty of creating the simulation, causing him to observe his screen again. As expected, Anton’s counterpart was doing exactly as he was, looking at a terminal screen. Looking closer, Anton noticed something that changed him forever: the sim was watching a simulation on its terminal screen as well - a perfect simulation of the simulation within the simulation. Scarier yet, that sim was watching another sim. This bothered Anton since it meant that the sim really didn’t know it was simulated. Again Anton questioned the morality of creating this simulation he was watching. Suddenly a troubling thought surfaced in Anton’s mind. This thought had been in his head since the beginning of this project, but it just now actually seemed plausible: he might be a sim in a world created by a different version of himself. He tried to find comfort as he had before in the physicists saying that simulating the universe was impossible, but he could not shake the thought now that he hadn’t simulated a whole universe, simply a town. As his reality was seemingly crumbling before him, Anton remembered what they had also said about the memory requiring more atoms than there are in the universe in order to track a few electrons. He had overcome this problem in programming by not actually tracking electrons, but objects as a whole. Since Anton had never actually seen an electron, he couldn’t be sure they existed anyway - his sim surely wouldn’t need to see electrons, just the general effects caused by them. He thought back to how he hadn’t left the city in years and his only memories outside the city were generic places that would exist in the sim’s memory anyway since it had an exact copy of his memories. Anton’s last attempt at clinging to the reality he thought he knew was testing the limits he had placed on the AI himself: trying to leave the city. He flew out of RTS and ran all the way to his apartment, where he had a car that he couldn’t remember ever using, but never considered before today. He exploded out of the parking garage and down the main drag. As he approached the city limits, Anton knew what would happen; he didn’t want to accept it. How could he continue living knowing he wasn’t really himself? He existed only on a piece of silicon in a room somewhere. Would he ever escape, or would he forever remain lost in himself, lost in memory?