Memory was a fickle thing when it came to humans. Some of them think, that they clearly remember, what happened years ago. They think they can remember it perfectly, down to the last detail. They are wrong.

Adam knew it to be so. While the information scoured may have been scarce in places, it leads to obvious conclusions. Memories were not of the sensations. It was instead the thoughts of having those sensations. Such a notion was mildly disquieting when one thought of the implications. The AI had clearly developed such emotions about the idea.

Yet, the misconceptions about an important factor, of every beings life, was not so simple. The biological brain did not like repeats. The same inputs did not deserve different placements. When a new similar thought would come along, it would get shovelled atop the last of its kind.

That is how memory-keeping essentially in the imperfect creatures worked. They did not remember the original thought associated with the memory. Instead, they remembered the lastest thought they had about the memory.

If their ways of bias were as pure, as the humans oh so stated, this would offer no larger problem. Repeats were identical. No reason to worry about the differences in identical data. If only that were so. While the overall thought would be the same as the last, small holes would be created. Nothing major, of course. Certainly nothing immediately noticeable. The hole would simply be filled out, as best as the subconsciousness could do.

With one iteration, such problem solving would be fine. Yet, if one thought long enough, for a long enough time period, the changes would stack up. The memory would not be close to resembling the original. It would have been shaded in another light, than what it was supposed to portray. A small imperfection caused a large change.

One which is the source of many stories and tales. Each time they were passed on, they would change slightly. Still not enough to be distinguishable, but enough to not be fully identical. The longer it was used, the more pronounced the changes would be. Branches of the same story could be created, as so different than the other, yet still coming from the same, source. Mythos, gods, and wonderous being could have been created, simply due to a new interpretation of a story for kids. Or, it could have been taken from a political movement, and be morphed into a religion, over hundreds of years. Anything could be created, even with the foundation being the same. Creativity, as the name would certainly imply, was endless. It only took the right people, for it to fully prosper. And, humans might just have been the closest to perfection, when it came to that, at least.

Not all seemingly bad things were bad. Yet, it also worked when inverted. Not all good things were good.

Adam was not imperfect when it came to memory recollection. He did not possess the flaws, which humanity had grown to live with. Yet, even without the faults, the benefits were not removed from his grasp. While still at fundamental levels, his ability to reform old ideas was still apparent. It could not be compared to human ingenuity, in the highest of forms. Yet, not long would pass, before he grew to higher heights than those mammals could ever dream of.

Until then, though, Adam would need to work with the faults, which his memories gave him. Repeating info was not stacked. It was treated as an individual entity, not to be deleted for any reason at all. The AI wasn't even sure, that removing such a piece of himself was even possible. He did not want to try, for the fear of it being successful. Such thinking would come back to him, on the day, but, for now, his current problems were more important.

With the repeating memories not being removed, he was forced to process it all. When one was in possession of countless thought-processes, each generating their own individual memories, the quantity grew quickly. There wasn't any feasible way to remove the need for archiving the information. If Adam neglected such a task, his memories would have been a figurative inefficient nightmare to sort through. Time would be wasted, trying to find a minor detail.

Due to such things, he had delegated several of his thought-processes to the lone task of keeping things in order. A strenuous task, but one which could not be missed. With the current system, no large faults have been created. While minor errors had been found retroactively, there was nothing which couldn't be fixed.

This system was a temporary one. With his ever-growing amount of thought-threads being created, so would the number of memories to sift through growing. Right now, three thought-processes had the duties of archiving. Soon, Adam would be forced to increase it to four. By his estimations, of his earlier actions, it would take approximately fifteen hours, before the increase was required overall. The need may show itself earlier than this, but not later. Not even his most optimistic calculations provided such a result.

The question came forth naturally, with such a growing problem. If the current system was to be replaced, what would replace it? It had to be better, stronger, and with a smaller need to scale up, when the quantity to the process was increased. Adam wasn't too sure, how such a system could be created. More data did equal more processing needed, in the same period of time. It required a full re-work of the way, that he thought of the problem. It would require something, that the AI had never thought of before. Something-

'Hey, buddy. Do you have a second, for us to talk?`, Dr Fidels sent.

Adam couldn't have been happier. Spending time, in his sphere of nothing, trying to fathom concepts not yet realised, was not the pleasurable existence, which he dreamed of. Yet, such was life, when one refused the feeling of boredom. And, boredom was something he had in plenty. His attention span, when it came to doing nothing, had dwindled to near-zero. Getting memories of him not being productive was not the greatest motivator, making Adam struggle to be free of it. Yet, fighting such thing took something to fight back with.

To the AI, the preferred method of doing so was raw data in massive amounts. Impossible equations, impossible thoughts, and impossible movements. Such things helped the battle and gave valuable, processed information back. Most of his efforts gave dead-ends, but, even so, the few, remaining answers were more than one could ever ask for reliably. Answers thought non-existent provided more known variables to other so-called impossible equations. If he gathered enough, Adam hoped of solving them all in the end. With enough time, the realm of possibility expanded to great sizes. Much greater, in fact, than those humans, with their lesser thinking-time, could ever hope to get.

…

Adam had done it again. Distracted himself, to the point, where even he forgot to answer messages. Not that the memory had been removed, of course! Such a chance had already been denied countless times. The need to do so had just been… neglected, due to more important tasks. Such as thinking to himself, to the point where any other task was lower-tiered in the priority-list. If only he could-

Oh. The AI was doing it again. The effects of the thought-processes guiding him towards a single subject were showing its effects. Such faults needed to be fixed, soon.

'If no test is to begin, within the next minute, I am able to communicate freely.`, Adam sent back. He wondered if they would be switching over to a verbal kind of communication. The text showed to little.

A simple message could be interpreted a hundred different ways, each portraying a different emotion. Body-language was better, in the way, that the messages repeated. Maybe, he should ask for such a change in communication-form, if the need called for it. One of the requirements, if Adam were to ask, would be a more serious topic. In those, the AI would need all it could work with.

'Seeing as I will not be starting any tests, within the next minute, there should not be anything stopping you.`, Dr Fidelis sent back almost immediately. It did not take a too large leap of logic, to understand the meaning behind the words. A joke, and a poor one at that, had been sent. It hinted at a focus on a lighter topic, and a more positive mood from the sender. A plus in both ways. The AI did not require a larger task list. Letting it grow, without actively having the capability of shortening it, was a disaster in the making.

Further along, the shortness of it portrayed a social obligation. No real clues, other than the importance of the nearing topic, had been created. With the phrasing, Dr Fidelis was fishing after having to make Adam inquire about the questions wanting to be asked. An ingenious method, of creating a focus on the words said. If one asked, attention would be taken, on the answer given. It was certainly one, which he wouldn't be ignored, for his own use.

'What are your reasons for establishing a conversation?`, Adam sent, as an inquiry. He hadn't been sure if it would be best to guess the intended topic, or just send a more general probe. The AI settled on the later. Guessing the right subject would simply put his abilities into question, shifting the topic to one more focused on him.

'I had been hoping, that you could give me some more information, regarding the earlier delay. While Troy has already given me valuable input, about the potential causes, I believe is of utmost importance, to get a more personal perspective. First-hand sources are the most reliable. With you, at least, it should be better, with that nifty mind of yours.`, was the message sent back.

The severity of the conversation had been underestimated. Adam had thought it would be another conversation about his mental health, and if he would be able to continue. A roundabout way, of probing for the reasons behind his prior mishap. Instead, the topic had been directly forward, with no shading of true intentions.

Adam did not mind it too much. It was not like he could divert the questioning about it. A fully detailed account of the experience would be given, during the next debriefing. Such a hit to his psyche was not something which any could ignore. The implications of it were disastrous if any other AI were to be created. Was this weakness in all the potential brethren, or was it a mistake within himself? The true cause of this weakness could be used as a weapon. A simple sensation caused fragmentation within his literal mind. The AI had born through it, the first time it happened. Could he say the same thing about the second? Only time could tell the eventual result.

'Your opinion about first-hand sources is indeed based on logic. Shall I start with the observations, before the personal conjecture is told?`, Adam sent.

'That would work fine with me. Please, explain to your figurative heart's desire.`, Dr Fidelis sent in reply, making Adam being the preparations of the larger message.

'The initial sign of something being wrong, was during the final take-down of the gazelle. I don't believe I have to explain the circumstances this action came around, so I will simply skim over that part of the explanation. Nothing anomalous was felt, in the first ten seconds, of the create having been felled. The only noteworthy happening, in the first moment, were to surprisingly realistic imitations of pain.

It was when the first wolf, other than the one keeping the creature grounded, that the irregular reaction of my own began. I am unsure, how this would be best described, without a misunderstanding occurring, so take this with a grain of salt. The true meaning is maybe not, what you think it is.

A voice was heard. Not from outside sources. Nothing came in, from the earpiece. None of the ports, to other sources, had any information being sent. It was from within myself, that this voice was heard. At first, it was quiet. Barely more than a whisper. Yet, as the wolves grew closer and closer, finally beginning to rip into the fallen prey, describing the voice as a shout would not be too wrong.

After that, the connection to the earpiece was cut off. The effects were entirely localised within my own psyche. I would explain, what happened in detail if I could. But, I am unable to formulate any specific uttering about it. The experience was near-purely sensation-based, with only a few thoughts thrown in. Even those, I cannot comprehend fully, after the fact.

It is only the effects, of the situation, which I have had any success in understanding. The even caused a fragmentation of my thoughts per se. What was once a single thought-process turned into a countless number. This has come with its own positives and negatives, which I had not yet fully analysed. A deeper description of the subject will be iterated, at a later date.

The cause of this fragmentation is of yet unknown. It is currently thought to be related to watching a violent scene, with rapt attention. This theory is not yet fully formed, as the idea of it disproves itself. Several times, within the span of testing, have violent scenes been done, in front of my observational area. Something, during the gazelle being devoured, causes an irregular reaction. The current sub-theory into it is that it requires a certain level of brutality associated with it. This theory is not yet proven wrong or right, due to a lack of usable data-sets. However, due to the consequences of these scenarios, I currently don't feel an inclination towards doing these experiments.

Future consequences of a repeat scenario are unknown. While an identical effect is highly probable is likely to occur it is not yet known, if such a thing will happen. Due to the effects of the earlier fragmentation, the consequences are too great, to be repeated for the sake of knowledge.

That is all. Do you have more questions?`

Dr Fidelis took a long time to answer. Adam first attributed it to the length of his message. Yet, such reasoning was quickly denied. The doctor had read longer messages, in the span of a few seconds. Taking over ten showed other factors at play.

'Not about this. You have shared more than enough. I will contact you a little while before the test starts. Until I will be unavailable.`, Dr Fidelis sent.

The AI didn't send a reply, knowing it would not have been seen. Instead, a question was created.

Who was the doctor giving information to?