On December 23rd, 2033; a massive explosion occurred at Artifact Labs, the leading company in AI development. The true cause of the explosion was never discovered and there were no known survivors. However, there were still many researchers unaccounted for, including the lead researcher Dr. Adams. The destruction of the lab erased all official data of the experiments being conducted, but employees not onsite at the time reported that Artifact Labs were attempting to develop an AI capable of independent thought. What follows is the journal of Dr. Adams from the start of the experiment dated 1/3/33 to the day before the incident on 12/22/33. This is the only know record of Project Eve.
***
1/3/33:
I've decided to journal alongside the development of Project Eve. The importance of this project demands that there should be some record of its developments that is written colloquially. While our official data will remain objective and unbiased, I feel it is my duty to the world to provide a glimpse to those behind it. The goal of Project Eve is not to simply create believable AI, but to create truly artificial life.
Eve will be an AI that can think for itself, make its own opinions, and become its own person. The project name was a poke at me by my colleagues. This project was something I had to push for and the comparison of it costing me a rib is there. This project will be my legacy, there is no doubt I am willing to sacrifice for it.
The principle behind Eve's AI will be that it learns organically. Everything will be broken down and viewed through the filter of the most basic learning principles; pain and pleasure. In other words, Eve will have sensors to mimic human senses and learn through association. Some stimuli will be seen as positive and others negative. Eve's AI seeks out these positive stimuli while trying to mitigate the negative stimuli. It seems simple, but it will get infinitely more complex as Eve learns to associate one stimulus with another into a web of association. Tomorrow we will begin setting up the instrumentation that will act as Eve's senses.
1/4/33:
As of this afternoon, all instruments are in place to start training the AI and have begun training the AI's foundation. This process is based on a manner of operant conditioning. For each of Eve's senses, we provided stimuli that are considered pleasurable to all living creatures. For example; for her sense of touch we provided temperatures that weren't too cold or too hot, all while forcing activation of Eve's "pleasure" coding. This way when her AI encounters new stimuli, she will compare it to her baseline and decide if she will like it or not. In turn, as more "pleasurable" associations are made along with her other senses, her baseline for one sense has the potential of changing over time.
The one setback we have hit so far is her sense of sight. My colleagues and I aren't exactly sure which stimulus would be best to baseline pleasurable. Some suggested simple shapes, others suggested pictures of peaceful scenes, but we hadn't come to a satisfying conclusion.
1/5/33:
I had a stroke of genius in the lab today. If we want Eve to be artificial life, her first sight should be the same as any other living creature's, meaning the first thing a creature sees for comfort is its parents. In a way, the rest of the staff and me are Eve's parents. We all gathered around when we turned Eve's eye sensors on. For the rest of the day, we trained her visual senses by either moving closer or farther away from her sensor while providing positive and negative stimuli. We plan to keep training all of Eve's foundational AI senses for the rest of the week. I will update the journal if there are any new developments.
1/10/33:
After returning from the weekend, we came back to a grim yet fascinating discovery. Data collected over the weekend showed Eve had a sort of panic attack, she consistently experienced negative stimuli at night time. After pondering why I believed I came up with a conclusion. Eve was afraid of the dark. My theory was that because we were always here in the daytime with the lights on she associated night with us being gone. With the lights being off all weekend long, Eve experienced more negative stimuli than normal.
It was Eve's first thought made by herself. While the implication is an exciting first step, I can't say I'm happy Eve's first thought was born from fear. After a short discussion we installed a nightlight in the lab.
1/11/33: