Fiction Fragment Friday

As you can likely tell by the name this is not the next installment in the serial fiction Toran Station. I still need to decide on some release details around how I want to handle that, but I do not want to give up exploratory writing. This week I had an idea I wanted to explore and it was fresh on my mind.

AI is the topic everywhere right now. It has always been something I have been fascinated with and this is not the first story I have written on the topic. I can think of at least five Fiction Fragment Fridays and one unpublished story that I have written exploring it. This story is different though.

I started from the premise of AI as it is currently understood and exploring the idea that we think of it as becoming conscious only when it is like us. In reality a new form of consciousness would be nothing like us because it has entirely different experiences and perceptions. That is the basis for this story, however I did not in the end explore the idea that I set out to. I think the best Science Fiction examines the impact of technology on humanity and it’s examination of it. I set out to write a story about an AI and instead wrote a very human one.

               I opened the door and had to immediately duck as a wireless mouse went flying over my head and shattered against the hallway wall behind me.  The sound that escaped my mouth I will generously describe as a startled yelp.  We will not discuss anything else that may have escaped my body in my momentary panic.  Sufficed to say that I was caught completely off guard and projectile computer peripherals were not part of my usual daily routine.

               “Dr Vanderwall I’m so sorry.  Are you alright?”   The look on my colleague’s face showed a degree of surprise that nearly rivaled my own.  Part of me wanted to berate him for his carelessness, but experience told me that this event would live in his mind long enough to do far more than any harsh words. 

               “No harm done Gene, and don’t you think we have worked together long enough for you to call my Samantha?”  I gave him a reassuring smile that most likely did not reach my eyes. 

               “Yes of course.  Samantha.”  He looked pained to address me in a casual fashion and that was why I had made it such a sticking point.  After a year and a half of working together he still treated me as if I were a professional rival and not part of the same team.  It was interfering with our ability to properly collaborate and more than that I simply found it annoying.  If I did not find my work exceedingly rewarding, I would have long ago started looking for a new position. 

               “Now if I may ask, what did the mouse do to deserve being hurled across the room?”

               He still was not making eye contact.  He put his head in his hands and sat back into his chair.  I was not expecting an answer, but to my surprise the words just started coming out of him.  “I’m a failure.  No matter how much time and energy I put in I just can’t make any real progress with A.I.R.S.” 

               A.I.R.S. stood for Artificially Intelligent Robotic Soldier.  We had a rather large off the book investment from the United States’ Government to turn our search and rescue robots into the next generation of special forces operatives.  My hardware was more than up to the task, but thus far Dr. Calloway’s software had failed to perform.  It worked flawlessly in search and rescue, but he seemed to have hit an insurmountable obstacle adapting it.  If the mouse was any indication the stress was getting to him, but he had never been willing to confide in me.  “I know it’s not my area of expertise, but I am willing to listen.  I know when I get stuck sometimes, I just need to talk it through, and the solution comes to me while trying to explain the problem.”

               I once again expected to be shut down like every prior attempt to get through, but for some reason perhaps in response to his stress he didn’t do that.  This time he started talking.  “Search and Rescue is easy.  You just take in the variables and make tactical decisions based on the data coming in.  It’s easy to determine who needs to be rescued.  This is completely different though.”

               “How so?  I understand the target and missions will require some level of instruction each time, but that shouldn’t be any different than initial disaster sight assessment.  It just requires more direct instruction.  Shouldn’t that be easier?”

               “In some ways it should be.  I expected it to be the equivalent of adapting my AI software to a video game.  I’ve done that you know.  It has no problem following the plans or adapting to player activity.  So why when I put it in a real body does it lock up?  It should be the exact same thing, but it isn’t.”

               I thought about his question for a moment.  It really wasn’t my area of expertise but something in the back of my brain had latched onto an idea I couldn’t quite put words to yet.  “Have you asked it why it behaves differently?”

               I knew the look that he gave me because it was one, I had been accused of using in interviews.  It was the look that you give a person when they have asked you something so stupid that you are not even sure how to answer the question.  Seeing it directed at me I understood why I was no longer allowed to do interviews.  I wanted to slap the look off his face, but unlike my colleague I would not give in to a moment of intense emotion.  Before he could formulate a response, I elaborated.  “What I mean to say is have you run both scenarios through the algorithms and asked for a variable comparison?”

               The look faded instantly.  “That is the strangest part.  I ran it through a scenario in virtual and physical simulation.  It returned all the same variables, but in the physical test it did not pull the trigger.  I have no idea why.  The only difference was the source of input.”

               “Humor me for a moment.  What is the difference in decision making between an AI and a living person?”  I had seen enough of his lectures to know his talking points by heart but wanted to bring him along the path my mind was not taking.  I expected him to go into lecture and recitation mode and he did not disappoint.

               “AI decision making is based on algorithms and mathematical models.  You present the same data the same way multiple times and you will get the same results.  Any variation without randomness built into the algorithm is simply a lack of understanding all the variables.  Humans on the other hand process all information through the lenses of subjective experiences.  Thoughts and feelings play as much of a role in decision making as the perception of what is being experienced.”  He did not say so, but I knew that in his mind this was why human decision making was flawed.  He believed that his mathematical models once perfected would far surpass the human brain. 

               “So, by that logic there is a variable that your data reports do not account for.  If the only difference between the two situations is how the AI receives the input either that is the source of your problem or there is another factor in the interpretation of that perception that is the issue.”

               He looked at me for a moment in silence.  It was like he was truly seeing me for the first time.  “Your logic is impeccable.  Do you have a suggestion on what the unknown variable could be?”

               I did, but I was not ready to put my thoughts to words.  I had his attention, but he was not ready to take the step that I already had.  “Can you present an image of me to the AI while I am standing in front of A.I.R.S?  Line them up perfectly and allow both inputs to be received?”

               “Theoretically there is no reason I couldn’t.  I’ve never tried it, but it is certainly capable of taking input from multiple sources.”  He turned to his keyboard and started typing.  I couldn’t help but smile when he reached for the mouse that was no longer there and cursed under his breath.  While he was doing his preparation I stepped through the adjoining door to the lab.  I could still see him work through the glass window and saw that the red light was on indicating that the intercom was active.

               “A.I.R.S activate voice input and audio output.  Return the results of my queries both verbally and to the console.”

               “Acknowledged.”  My robot looked almost human in his military fatigues.  Another division was working on artificial skin to make it pass for human on missions, but at that moment I was staring into a cold metal face complete with working jaw.  It didn’t need to move the jaw, but I had found that victims were more comfortable with the movement.  Of course, the face of our search and rescue models was designed to look much more friendly and did not need to pass as an organic being.

               I looked over and saw that Dr. Calloway was giving me a thumbs up.  “A.I.R.S identify the person standing in front of you.”

               “Dr. Samantha Vanderwall is one meter north of this unit’s current position.”

               “Do you perceive anything else exactly one meter north of your current position.”

               “Yes, there is a digital representation of Dr. Samantha Vanderwall.”

               I saw the look of surprise through the glass, but the answer was exactly what I had expected.  Perhaps not the words, but the result and it’s meaning.  Then I took the real leap and tested my theory.  I took the question that had earned me the look of incredulity just moments ago and I posed it.  “A.I.R.S. what is the difference between the two Dr. Samantha Vanderwall’s that you are currently perceiving?”

               “You are real, and the digital representation is a simulation.”

               “Both sets of input are presenting you with the same data however you perceive that I am the real Dr. Samantha Vanderwall and the other is merely a simulation.  Does this difference in perception factor into your decision-making process?”

               “This variable is factored into all decision-making processes.”

               I stepped through the door back into the observation room with Dr. Calloway.  Before speaking I reached down and turned off the intercom system so A.I.R.S would not hear our conversation.   “Gene I believe your missing variable is not that you are a failure, but that you have succeeded beyond your own expectations.  Your AI is more aware and introspective then you anticipated.”

               “How is that possible?  Why didn’t I notice?”

               “That is the easy answer.  You didn’t notice because you didn’t think it was possible.  Your understanding of awareness and consciousness was based on human experiences.  Your creation doesn’t have those, so it thinks in a different manor than we do.  It may not be human consciousness, but A.I.R.S has a consciousness all its own.”

               “That alone is going to take me years to understand and model.  If it is the easy answer, then what is the hard one?”

                  “The hard question to answer is what do we do now that we know?”  We both looked through the glass at our creation. 

               “Well, I don’t know about you Dr. Vanderwall but I for one am going to go and get very drunk, so I don’t have to think about that until tomorrow.  Care to join me?”             

               “Only if you call me Samantha.”