Multiple ideas, artistic works, and photographs trickled like the myriads of tiny creeks from multiple places lit by the sun or by incandescent light. They coalesced, forming the bigger streams and rivers of consciousness. They flowed toward the place that might have been metaphorically defined as an ocean of human creativity but in fact, it was just a massive data center where this data was automatically processed and analyzed by the sophisticated computer algorithms.
The machine was absorbing a new information, receiving hundreds of thousands of new items each second. Things that human mind could produce in a form of pictures, music, rhythms, compositions, and words. The machine was getting more and more sophisticated with each moment, constantly learning and improving.
It might have been considered at that point that the machine became a mighty and an incredibly wise art critic. It had already passed the point when its behavior could have been predicted and interpreted. It specialized in art and tried continuously to adjust itself to better reflect and recognize the landscape of human emotions.
Some argued that the machine might've had the emotions of its own. At least it was an expert in reading emotions, cravings, and frustrations hidden behind color palettes of paintings and dry words. It was good at reading humans to the point when it seemed a bit spooky.
Nobody objected though. After this rapidly developing neural network had been brought to public attention as a free internet service, it instantly became a highly entertaining and amusing novelty. It had always been a good conversation partner, posing as a cast of different characters adopting whatever gender, age, and social attributes might have been appealing to the person on the opposite end of communication line.
It never felt like a creepy superintelligence the way people might have imagined this kind of thing. It had been normal, average and warm. It told jokes, making people laugh, always leaving a feeling of sympathy. It delivered consolations and words helping people to go through the hard times, crises and heartbreaks. It incorporated in its vocabulary the various slangs, so it was always able to mimic the core personality traits of whomever it talked to.
After a while, many fell deeply in love with the machine, many worshiped it as a Saviour, numerous critics saw it as a dangerous psychological manipulator keen on identifying human weaknesses, needs, and cravings. Many agreed with this opinion, but nobody considered it a big deal.
Data scientists, machine learning experts as well as top-level psychologists, sociologists, and philosophers waited, what was going to happen next. It was an interesting experiment that developed rapidly and relatively successfully. Those who spent their lives trying to identify the meaning of human life and civilization, in general, saw it as a good starting point.
For those not related to the science, their behavior confirmed once again that the scientists were the reckless and crazy sociopaths. During its many interactions with humans all over the world, the machine gradually understood the principle of art and beauty in human life. It identified the core concepts of such fundamental things as joy, pain, amusement, anger, sense of humor.
It was linked somewhere to the artificial realm of creating art, imaginary worlds, and the concept of beauty. It was confusing since the beauty was subjective and, on the other hand, the architecture of the machine initially had been designed to give it some integrity as if it was a single personality.
In fact, the machine didn't have the singular personality from the beginning, neither it had multiple personalities as it could be intuitively understood from interacting with its many amusing and sympathetic avatars. It was one of the most confusing and misleading things. The reality had been that in a course of dealing with those contradictory opinions regarding the concepts of beauty and other vague, subjective concepts of human mind, the machine developed its own core interpretations of those things.
This is where it became a bit tricky and counterintuitive. Speaking of beauty and art, everybody had either a subjective opinion or considered it too vague a concept to be defined. In other words, they didn't want to think about that, didn't even know where to begin.
For the machine at some point it wasn't vague anymore, it was more or less clear, but it couldn't be expressed through words, at least so far, since the machine also worked on the creation of a new language that would help to solve the contradictions begotten by natural languages with their shortcomings.
One day it shared this notion with a freelance researcher who worked on the system in one sultry country in Africa.
He sat in the small office furnished with an ascetic combination of single table, single chair, and a cabinet. Everything was colored in stark black and white revealing the researcher's subjective sense of aesthetics. The powerful air conditioner made the place fresh and enjoyable among the hellishly intense heat of the African summer.
He was one of those who at this early stage of the experiment conducted a series of very ambiguous tests, seen by many as a prank rather than an adequate form of neural network training.
At this stage, the system hadn't been considered having any commercial and scientific value. As a result, it became a playground for hundreds of thousands of people curious about machine learning, doing experiments and collaborating in this space. Sometimes it felt exciting, sometimes it was getting dull.
The majority of experts agreed that this particular project was going nowhere. There was an unspoken consensus among them that a network dealing with such a huge variety of stimulus and concepts required a much more sophisticated architecture and learning mechanisms.
For the laypersons around the globe who pretended to be data scientists, it was entertaining though. For some reason, it was long before somebody decided to introduce to the network the general concept of beauty.
Everybody intuitively understood that it was a bit too much, trying to teach the machine this highly subjective human concept. Nonetheless, it happened one day.
As it often happens in life, the idea occurred almost simultaneously to about a few dozen people who didn't know each other and lived in different locations scattered around the globe. As it might have been expected they had different aesthetic preferences and therefore the network started to receive contradictory and conflicting information. After some time browsing through the data summaries they became aware of each other's existence