Protocol: f145c0

Day 1: Testing in the lab. We have been experimenting on a new automaton model and it seems to be working well. We are able to teach it basic actions and it performs them reletively effortlessly. Building with blocks, sweeping up floors, inputting sequences on a control pad, and a lot of other simple tasks.

Day 12: Things are still looking good. The Android has learned full patterns of schedule and will remember each day what we told it to do the previous day. Adding on more tasks and switching things around is what we are going to try next.

Day 25: It has been going swimmingly. The Android adjusts smoothly to sudden changes in schedule. It understands our commands and reacts in real time at lightning speeds. It's time to begin testing it to accompany an agent in a simulated scenario.

Day 30: Simulated scenarios are going well. The Android stands by, waiting for instruction from its agent partmer. We need to tell it something and wait a few days with it in an idle room alone to see if it still remembers commands after that.

Day 33: The Android actually remembered its commands! We left it in a room with no communication for about two days and it seemed to behave normally, standing and waiting for one of us to give it instructions. It seemed to go into a sleep mode when not receiving input for quite some time, which is normal, according to the model guidebook. After waking, the Android immediately took to the agent's side and performed all the tasks that had been assigned prior without fault. It is time to test the Android's own reflexes and simulated judgement.

Day 38: It is taking the Android some time to act on its own without having to perform what we tell it to word-for-word. It seems to sit there awaiting commands. We try to give it a couple of choices, but it refuses to do anything until we give it explicit guidance it on which option to pick.

Day 43: Finally making a bit of progress. We had to work around by giving the Android a command to do "something". When told to do "something", the Android will now perform a random task. This is one step closer to allowing the Android its own agency in simulated judgement.

Day 45: The Android is now able to choose things. If we give it two options, we can command it to pick an option and it will. Unfortunately, it all seems random. There is hardly any solid reasoning behind the choices it makes, but we will continue working with it.

Day 60: We are trying to communicate with it more and get it to tell us the why behind the what.

Day 67: oh no. I don't know what happened! We walked into the lab and found that the Android was not in its spot. It has escaped and has hidden somewhere in headquarters. We tried to check the surveilance cameras and it seems to have acted completely on its own. We need to find it before something horrible happens.

Day 80: After an unfortunate turn of events, we have declared a Protocol: f145c0.