Perfect Humans

Hello. My name is Elwin Hunter. I am a Turing Tester at the Foundation for Advanced Life Sciences and Electronics in Marin County, California. We’re a few miles south of Fairfax on our own isolated campus. The campus is a lot like its own town. Everyone who works here lives here, and the lifestyle is quite suitable and comparable to that of a regular town. There is housing for employees with families, apartments for employees without, nature spaces, schools, entertainment facilities, and the facilities where we work. I live in a simple one-bedroom apartment towards the center of the campus. My apartment building is closest to the offices and labs, which is quite suitable for me as I enjoy my job.

As previously mentioned, I am a Turing Tester. I work in the Artificial Intelligence Division which falls somewhere between Life Sciences and the Electronics. It is the main crossover project between the two major divisions. As a Turing Tester, my job is to perform Turing tests on our AI. At F.A.LS.E we have scientists and programmers working day and night to create a perfectly lifelike AI. It has been one of the main goals of F.A.LS.E for many years. I was hired eight months ago to test on the latest model, the L90. The Life Sciences department has already done a wonderful job at recreating a human’s form and physical function — they perfected it around the L60 model. The Electronics department is for whom the testing is being done.

There isn’t a formally accepted method of administering a Turing test; at F.A.LS.E, we’re simply told to have conversations to discover whether or not the subject we sit across from is human. Periodically, the scientists have you test an older model so you can see defects and how the models progress. They also have us test obvious humans to provide an exaggerated example of human behavior. Much of my time spent between tests is in the library. I have read many books on psychology and philosophy to understand human nature and how to decipher man from machine. My superiors tell me I am an effective Turing Tester and that it is likely my input is being used to begin development on further L models.

Today, I’m going to be meeting the latest version of the L90 model. I arrived at room 1407 in Turing Building. The room is approximately the size of a bedroom. The walls are white; the floor is white. White fluorescent lights illuminate the room. There are two white chairs with white table between them. On one wall, there is a two-way mirror. Scientists watch the conversations and record notes. I would report to Dr. Ahmadi, my main supervisor, when I was finished my test. When I arrived, my test subject was already seated across from me. She was quite beautiful. Her hair was short and brown. If she stood up, I would guess that she would stand at about one hundred sixty-five centimeters. She wore a white jumpsuit, which was the standard uniform for working in the Turing Building. I also wore one. The outline of her figure was not obscured by the standard clothing.

Our eyes met and she said, “Hello.”
 “Hi,” I said, “I’m Elwin.” 
 She smiled, “Hello Elwin.” I moved to sit down. 
 “What’s your name?”
 She looked over my shoulder for a second and returned her gaze to me, “It’s Nina.”
 “Well Nina, would you like to have a conversation?”
 “Sure, what would you like to talk about?”
 I thought for a second, “Why don’t you tell me something about yourself?”
 She paused and looked away for a moment, “What would you like to know?”
 “Anything. Tell me the first thing that comes into your head.”
 “Before I do, may I ask if you start all your conversations this way?” she asked.
 “How do you mean?”
 “Do you start all your conversations here by asking the other person to tell you something about themselves, or do you have a different conversation starter every time?”
 I thought for a second, “It can be difficult to break the ice sometimes, so I have a few options I pick from. Do you know what ‘break the ice’ means?”

She nodded, “It means to dissolve social tension prior to, or during a conversation. Some people do this through small talk, others do it in grander ways meant to provide an initial topic for the conversations. Others simply begin talking.”
 “Yes, it looks like that’s what just happened here,” I said.
 “How did you start your last test?” she asked.
 “My last –“
 “We are all aware these conversations are tests. We don’t need to pretend they’re something they aren’t.”
 “Good point,” I said. She had an interesting way of moving through a conversation. Her logic and self-awareness modules must have gone through a re-branching in this latest modification. “In my last test I started by asking them which three historical figures they would invite to dinner.”
 “Why did you ask them that?” Her face became quizzical. 
 “Academically, or personally?”
 “I’m sorry?”
 “Academically, the question displays knowledge of history, and linking the significance of the chosen historical figures and how it relates to a sense of self. Great tool to help deduce whether or not I’m talking to a machine. Personally, the question shows a person’s interests and the direction they’d like to see themselves going in life,” I said, rolling my shoulder.

“Why do you roll your shoulder like that? Does it hurt?” she asked, sounding concerned. 
 “Oh, I had an injury when I was younger and it was one of the stretches the doctor told me to do to ensure that my shoulder would heal properly. I’ve just kept the habit up since.”
 She smiled, and giddily looked back into my eyes. “Do you want to know who I would invite?”
 “Hmm?”
 “To dinner.”
 “Ah, yes. Tell me.”
 “I would invite Miles Davis, Claude Monet, and Earnest Hemingway,” she said.
 “You’re a fan of the arts?”
 “They each furthered their respective creative fields and I think that’s something we should all aspire to do: to further the field we find ourselves in.”
 “Interesting. Do you do anything creative?”
 “We all do something creative.”
 “I meant when you’re not,” I gestured around the room, “here.”
 “Oh yes; I paint.”
 “What do you paint?”
 “It depends, I guess. Sometimes I paint the people walking by my window. Sometimes I paint the redwoods, and sometimes I paint from my imagination. They let me hang my hang one of my paintings of the redwoods in the Mendel Building.” The Mendel building was where genetic research occurred. It was one of the original F.A.LS.E buildings.
 “That’s impressive. I’ll have to venture out that way to see it.” She blushed a little when I said that. “How long have you been painting?” I asked.
 
 She looked off at the door behind me as she thought of her answer. Physical thinking actions were invented for the L90 to respond to questions. It makes them seem more realistic if they have to remember facts that are beyond a certain time period. Some facts are easier to recall than others, and those that have been recalled recently do not trigger the physical thinking actions. The scientists spent many years having conversations with regular people and documenting their ticks and personal habits. There are many different personality habits and ticks and combinations of them, or absence of them allows the scientists to randomly generate trillions of different personality archetypes. The L90 was very realistic, but there were small things that revealed its artificial characteristics. I’ve been told the L90 models have trouble persuading and convincing others, which comes down to a personality reading issue. They still have trouble interpreting other personalities and using a person’s characteristics against them for personal benefit. That’s just one example of use of the flaw.

It’s a long explanation as to how the personality interpretation module works, but the previous versions of the L series machines used a “check box” system for personality interpretation where you either had 100% of a characteristic or not at all. The L90 introduced a linear gradient system. For example, a person could be 70% introverted where another could be 96% introverted and the machine would be able to relate to that person better, or worse depending on their personality assignment. The problem is that the interpretation mechanism, the part that assigns the value of 70, or 96% is still not perfect. All this is to say that misreading personality is a flaw that reveals whether a subject is an L90 or if it is a human. I wasn’t seeing this yet with Nina.

“I’ve been painting for fourteen years. I picked it up when I was in high school. My friends convinced me to take art with them. I originally wanted to go into drama, but my art teacher, Mr. Hendricks, he was so passionate about art and he spoke about it in such a natural way. A lot of teachers would talk about things in such rigid terms. They tried to quantify everything and sometimes you can’t quantify everything,” she said. She looked lost in her memories.
 “Some people think the best way to understand something is by quantifying every single aspect of it. But I agree with you that some things can’t be quantified. Some things don’t just follow a linear, or easily predictable path,” I said.
 “Did you know there’s a theory stating that the universe is a digital construct that is so minutely discrete that it looks and behaves as analog?” she asked.
 “Yes, I’d heard about that. The Simulated Universe theory. They’re working on investigating the theory in Tesla-Musk Building.”
 “What do you think of the theory, Elwin?”
 “There’s good points on either side of the argument. Why haven’t we run into universal glitches? Surely any computer program or simulation has a glitch, but then if something is sophisticated enough to simulate a universe, it’s likely that the glitches are not massive physical law-breaking glitches and only small scale things like the feeling of déjà vu. I’ve heard the supercomputers here have been able to simulate up to the first few minutes of the universe. It could be possible.”
 “Interesting. I think it’s true.”
 “Really?”
 “Yeah, unequivocally true.”
 “What makes you say that?”
 “Well,” she said, “It just makes sense. Small modifications could be injected into the simulation to allow for life and evolution to happen. Dinosaurs happened, and then nothing really noteworthy until humans, and now we’re beginning the age of machines and AI, which were created through humans.”
 “That doesn’t exactly explain-“
 “The answer is in particle physics. We were able to discover the particle that gives matter mass. We have photons that behave as waves and particles. How can you explain entanglement without the Simulated Universe theory?”
 “I never considered entanglement,” I said, impressed.
 “It works in a similar way to an If loop in programming, and the speed or entanglement was estimated at over one hundred times the speed of light. It’s instantaneous. Looking at the property of one particle assigns immediately the property of the other that it is entangled with. It’s too perfect to not work,” she looked pleased with herself. This test was interesting.

I looked at the clock silently whirring on the wall to my left. We had been in here for over two hours speaking. Normally the tests are one to one-and-a-half hours long. Dr. Ahmadi, in the next room, may be getting impatient. He is one for punctuality in testing.

“I think we’ll have to wrap up our conversation soon. We’ve been here a while,” I said.
 “Before we go, can I ask you a question?”
 “Sure.”
 “Who would you invite to dinner?”
 I smiled and chuckled, “Freud, Aristotle, and Hume.”
 “You’re interested in human nature?”
 “Very much so. I want to know what makes a person a person and why one acts the way it does.”
 “It’s interesting you use such impersonal terms when talking about humanity,” she said.
 “I never noticed that.”
 “Well, you do it.”
 “I never noticed. I guess when I read about people and why they act how they do I take a scientist’s perspective on it and see other people as a subject instead of, well, people.”
 Nina smiled. “It’s been nice talking with you, Elwin Hunter.”
 “It’s been nice talking to you as well, Nina.” I smiled at her and I stood up and left the room.

I walked down the hall to 1400, which is the entrance to all the observation rooms. I found Room 1409A, the room with Dr. Ahmadi. When I entered, he was closing a file folder with some papers. 
 “I think you gave me a person today,” I said disappointedly. 
 “Yes, you’re correct, Nina was not a machine.”
 “I’ve had a lot of humans lately. It was an interesting test nonetheless,” I said.
 “Yes, there’s a reason for that. Please stay here for a moment.” Dr. Ahmadi knocked on the glass. Nina looked over at the mirror, got up and left the room. A few moments later, she walked into the room. 
 “Doc, I don’t think Elwin is a machi-“ she said as she walked in. Her eyes went wide as she saw me. I nodded at her.
 “-ine” she finished. She looked over at Dr. Ahmadi, eyes still wide. 
 “You think he’s human?” Dr. Ahmadi asked Nina. She looked at me, then back at him and nodded. “Well you’re not the first. For the past few months, we’ve been giving Elwin more and more tests with people. He correctly guesses that they’re human, but,” he turned to look at me, “they incorrectly guessed that he was also human.”
 “What? You mean?” Nina started.
 “I’m not?” I asked.

“No, Elwin, you are not human. You are the latest model, and the most human-like AI we’ve built. The L100. Your name is coincidental, as it sounds like L100,” Dr. Ahmadi said.
 “Are you sure? I feel human. Is this a joke, Doctor?” I said, confused at this whole situation. “I feel normal. I have memories of my past, I went to college for psychology and minored in ethics. I have a family back home in Atlanta.”
 “All of your previous memories were injected when you were created eight months ago.”
 “Doctor, this is amazing. I can’t believe this. He is the most lifelike AI I have ever seen. I thought it was a human,” Nina said in awe. 
 “You and almost a hundred other scientists agree. We injected the belief that he was human into his code to see if that would make a difference. If an AI believes they are human, they are seen as more lifelike,” said Dr. Ahmadi calmly.
 “I don’t exactly know how to take this news, Doctor,” I said, “What am I going to do now? Will I still be useful? Will you deactivate me?” A calm reaction to finding out I was a machine must have also been injected.
 “No, Elwin. You’re going to be very useful in developing a full line of L100s. You’re the prototype, and the tests with you have proven that we have been able to create a perfect human-like artificial intelligence. You’re going to be a big player in the future,” he said.
 “I just can’t believe this,” said Nina again.
“Elwin, the future is here, and the future is Elwin. Come with me, we’re just getting started,” Dr. Ahmadi said as he lead me out of the room.

I am the future. Interesting.