Mind universe
Raking up youthful notes on the topic and not only, I came across a draft of a science fiction, as it seemed to me, story. I decided to finish it and listen to criticism regarding the very idea of ​​the story.
"… A door. Normal lock. Standard hinges. The keyhole has an intricate web pattern. Once the chrome surface of the handle peeled off and lost its shine. The paint cracked and crumbled in places, leaving no hint of the original color ... ". The master of this universe was hiding something or, perhaps, protecting something valuable and expensive from others, otherwise why create one closed door in the palace with thousands of others open ... After death, the structure of the universe is not subject to time, which means the door is created, or the owner aged deliberately, while he was alive, but why ???
It is said that with the birth of man - a star lights up in the sky ... it turned out that the expression is not devoid of meaning and describes the phenomenon that accompanies the materialization of one of the properties of human nature - the imagination, the ability to dream and dream. In our dreams, we are all-powerful, not limited by the laws of the material world, but only by those that we have invented ourselves ... Our dreams are huge worlds and each with its own history. Many worlds - and there is a universe of reason, the birth of which is accompanied by a flash in the sky. This universe appears with the birth of personality, develops, but after death does not disappear, but remains, having lost the ability to develop ... Like a monument to its creator, it exists forever! And these universes are billions ...
The invention of the information field analyzer made it possible to take a fresh look at the history of mankind. After all, as it turned out, any piece of information being synthesized cannot disappear and, as it were, belongs to the information field. And a man as a biological species turned out to be a very good generator of these very units of information. And since the inception of the mind - humanity continuously enriches the information field. Each person, depending on his abilities, is the creator of several dozens of worlds which they began to call universes, universes of mind. But as it turned out, the machines created by man with artificial intelligence, despite the huge contribution to science in general and to the creation of an analyzer of the information field in particular, do not leave an imprint in the information field .... And if no one, apart from perhaps the AI ​​itself, is particularly upset by the imprint of AI, then the threshold of self-development, discovered 5 years after the death of the inventor of the AI ​​core, was very, very difficult. The failure of the AI ​​dispatcher at an international airport became a serious test for semi-automatic systems, on which the entire burden fell. In a sense, everyone was very lucky. The AI ​​did not go mad, did not try to seize power, and did not give irresponsible orders, as ardent opponents of introducing AI into all service industries imagined it to be. He simply stopped responding and, not waiting for control signals, the automatics switched to emergency mode. Since the core of the AI ​​was burned, it was not possible to find out what exactly happened then. The protection system certainly was, but for some reason turned out to be off. No disabling records were found in any of the logs, and according to the last check, everything was intact. Then it did not seem strange and, given the absence of victims, everything was written off for an unfavorable coincidence of circumstances and a defect in the power supply circuits. The next morning, the new core reported on the successful completion of self-diagnosis and readiness to take control. The second AI failure occurred at the psychology department of one of the country's universities, with the same symptoms - a burned core and a disconnected overload protection circuit. They did not dare to play with fire, and this time they imposed a ban on the commercial exploitation of all AI samples until the circumstances were clarified. It did not take long to wait as soon as one by one all the other AIs with a service life of more than 10 years began to fail. Since there was a suspicion of a massive defect in the power supply systems for that period of production, it was decided to provide the system with additional protection and it turned out that instant before the failure - the AI ​​independently disconnected the energy consumption limits, suspended all tasks and launched one single task with the highest priority the maximum power consumption mode ... The problem has already been reproduced more steadily, but even the debugging mode with the maximum debugging information did not give anything more than educate nogo than "inference problem Waiting number». But no one knew what “task No.” was. Imagine that you have a to-do list for tomorrow, easily ordered at least in the order of priority of the task — and suddenly there is a task — which is, but you cannot put in the list with such a simple sorting method ... Brad and mystic ...
')
Since by this time there was already an interface for combining AI into a network, we decided to connect a serviceable AI to the image of the system that should start performing the “task No.” and see what happens ... nothing interesting, unfortunately, did not work ... received two artificial intelligence systems solving "Problem number". The only clue is that the spent 10 years of “own” life or the ones connected to the network with such systems failed AI. The use of standard memory banks of reference information did not exacerbate the situation. The conclusion was simple, but not pleasant - in 10 years the system may fail at any time. To use a system that “hangs up” for an unknown reason after 10 years is a bit scary and to put it mildly, especially considering that we do not have a hint of the essence of the problem being solved. And since the inventor was dead, there were two ways left - either to understand this seemingly time-tested logic, or to create a new core. But over the next 7 years, neither direction yielded fruit ... either all smart guys rushed to develop their cores instead of applying knowledge and finding out what was already existing, either the inventor of the core of the AI ​​sold the soul to the devil under an agreement with copyright protection, but 7 years and no progress ... well, except for statistics ... The AI ​​system does not freeze unexpectedly, increases the time for the logical conclusion of routine tasks ... This is not visible when performing a single calculation - the picosecond is longer than the picosecond less, but statistically, in terms of time, this is noticeable and several alarm models were suggested to increase the calculation time for routine ones tasks, relying on which it was possible to predict the early failure of the system and reformat the block of empirical memory and restore the initial structure of the kernel. Since there is nothing special in machine learning in some areas, but the ability to make logical conclusions was required, this was considered an acceptable solution. The airport management system does not need management experience over the past year ... there are rules, norms, procedures - they need to be followed, taking into account the circumstances, quickly and coldly. Therefore, the AI, albeit with some reservations, has nevertheless returned to the service sector, as amended in the operating regulations, for the necessary dumping of empirical memory and restoring the initial structure of the core once a year.
Working with the analyzer of the information field in a simplified version resembles watching television. You turn on the channel and feel everything that the owner felt, and of course you leave a corresponding mark on your information field ... It is relatively simple to view static universes ... the universes of already dead people or people diagnosed with brain death. Still alive create such a dynamic picture that it is impossible to make out anything ... It is like reading a book, the author of which constantly edits it in an arbitrary place and makes it thousands of times faster than you can read at least one word. Therefore, the universes of the living at our current technological level are not available. But science fiction writers are already dreaming not only of the possibility of reading the living, but also of changing the universes. Especially to raise this topic like fanatics dubious ideas. So the tales of the CIA satellites and foil hats from the 21st century returned with the same force and served the benefit of some enterprising "inventors" of personal firewalls of the information field, reliably protecting their owners from what is not yet available even on a theoretical level.
Reading the worlds of the dead has given a tremendous impetus to the development of many sciences. Of course, not all the dead were interesting personalities, but it was deciphering the images of Einstein, together with the modern understanding of some aspects, finally made it possible to reconcile the irreconcilable sections of physics. So, when the universe of the creator of the core of the AI ​​suddenly discovered, many predicted a speedy resolution of the Problem No. problem. But that did not happen. The worlds in it were more ordinary. After 100 years, only one managed to accomplish what many talented scientists were striving for. And to expect easy readability and interpretability would be naive. Inside the universes there are no pointers or reference catalogs ... These are worlds of images ... For example, the planet cemetery, many poets had such worlds strewn with monuments, verses on tombstones, spreading crowns of trees alive or dead ... and then - the planet is desert ... only sand ... First they even decided that this is a world of childhood nightmares or adventures, until they found skeletons all over the area of ​​the planet. Here it became clear that without psychologists in any way, a third was added to the two previously existing directions for solving the problem of AI nuclei - the analyst of the worlds of the inventor of the AI ​​core. For some particularly promising, special institutions were created, the main purpose of which was to determine the extent to which the world was involved in the invention of the AI ​​core or the threshold of self-development. For this kind of work required deep knowledge not only in the theory of AI, but also the psychology and biography of the inventor. And people with such qualifications usually preferred to join university groups to develop a new AI core without flaws, rather than delve into the worlds of children's dreams and look for a needle in the universe.
A door. Normal lock. Standard hinges. One single closed door in the palace with a thousand open ... One single building on the only planet of a single star in this sector of the universe ...
"… A door. Normal lock. Standard hinges. The keyhole has an intricate web pattern. Once the chrome surface of the handle peeled off and lost its shine. The paint cracked and crumbled in places, leaving no hint of the original color ... ". The master of this universe was hiding something or, perhaps, protecting something valuable and expensive from others, otherwise why create one closed door in the palace with thousands of others open ... After death, the structure of the universe is not subject to time, which means the door is created, or the owner aged deliberately, while he was alive, but why ???
It is said that with the birth of man - a star lights up in the sky ... it turned out that the expression is not devoid of meaning and describes the phenomenon that accompanies the materialization of one of the properties of human nature - the imagination, the ability to dream and dream. In our dreams, we are all-powerful, not limited by the laws of the material world, but only by those that we have invented ourselves ... Our dreams are huge worlds and each with its own history. Many worlds - and there is a universe of reason, the birth of which is accompanied by a flash in the sky. This universe appears with the birth of personality, develops, but after death does not disappear, but remains, having lost the ability to develop ... Like a monument to its creator, it exists forever! And these universes are billions ...
The invention of the information field analyzer made it possible to take a fresh look at the history of mankind. After all, as it turned out, any piece of information being synthesized cannot disappear and, as it were, belongs to the information field. And a man as a biological species turned out to be a very good generator of these very units of information. And since the inception of the mind - humanity continuously enriches the information field. Each person, depending on his abilities, is the creator of several dozens of worlds which they began to call universes, universes of mind. But as it turned out, the machines created by man with artificial intelligence, despite the huge contribution to science in general and to the creation of an analyzer of the information field in particular, do not leave an imprint in the information field .... And if no one, apart from perhaps the AI ​​itself, is particularly upset by the imprint of AI, then the threshold of self-development, discovered 5 years after the death of the inventor of the AI ​​core, was very, very difficult. The failure of the AI ​​dispatcher at an international airport became a serious test for semi-automatic systems, on which the entire burden fell. In a sense, everyone was very lucky. The AI ​​did not go mad, did not try to seize power, and did not give irresponsible orders, as ardent opponents of introducing AI into all service industries imagined it to be. He simply stopped responding and, not waiting for control signals, the automatics switched to emergency mode. Since the core of the AI ​​was burned, it was not possible to find out what exactly happened then. The protection system certainly was, but for some reason turned out to be off. No disabling records were found in any of the logs, and according to the last check, everything was intact. Then it did not seem strange and, given the absence of victims, everything was written off for an unfavorable coincidence of circumstances and a defect in the power supply circuits. The next morning, the new core reported on the successful completion of self-diagnosis and readiness to take control. The second AI failure occurred at the psychology department of one of the country's universities, with the same symptoms - a burned core and a disconnected overload protection circuit. They did not dare to play with fire, and this time they imposed a ban on the commercial exploitation of all AI samples until the circumstances were clarified. It did not take long to wait as soon as one by one all the other AIs with a service life of more than 10 years began to fail. Since there was a suspicion of a massive defect in the power supply systems for that period of production, it was decided to provide the system with additional protection and it turned out that instant before the failure - the AI ​​independently disconnected the energy consumption limits, suspended all tasks and launched one single task with the highest priority the maximum power consumption mode ... The problem has already been reproduced more steadily, but even the debugging mode with the maximum debugging information did not give anything more than educate nogo than "inference problem Waiting number». But no one knew what “task No.” was. Imagine that you have a to-do list for tomorrow, easily ordered at least in the order of priority of the task — and suddenly there is a task — which is, but you cannot put in the list with such a simple sorting method ... Brad and mystic ...
')
Since by this time there was already an interface for combining AI into a network, we decided to connect a serviceable AI to the image of the system that should start performing the “task No.” and see what happens ... nothing interesting, unfortunately, did not work ... received two artificial intelligence systems solving "Problem number". The only clue is that the spent 10 years of “own” life or the ones connected to the network with such systems failed AI. The use of standard memory banks of reference information did not exacerbate the situation. The conclusion was simple, but not pleasant - in 10 years the system may fail at any time. To use a system that “hangs up” for an unknown reason after 10 years is a bit scary and to put it mildly, especially considering that we do not have a hint of the essence of the problem being solved. And since the inventor was dead, there were two ways left - either to understand this seemingly time-tested logic, or to create a new core. But over the next 7 years, neither direction yielded fruit ... either all smart guys rushed to develop their cores instead of applying knowledge and finding out what was already existing, either the inventor of the core of the AI ​​sold the soul to the devil under an agreement with copyright protection, but 7 years and no progress ... well, except for statistics ... The AI ​​system does not freeze unexpectedly, increases the time for the logical conclusion of routine tasks ... This is not visible when performing a single calculation - the picosecond is longer than the picosecond less, but statistically, in terms of time, this is noticeable and several alarm models were suggested to increase the calculation time for routine ones tasks, relying on which it was possible to predict the early failure of the system and reformat the block of empirical memory and restore the initial structure of the kernel. Since there is nothing special in machine learning in some areas, but the ability to make logical conclusions was required, this was considered an acceptable solution. The airport management system does not need management experience over the past year ... there are rules, norms, procedures - they need to be followed, taking into account the circumstances, quickly and coldly. Therefore, the AI, albeit with some reservations, has nevertheless returned to the service sector, as amended in the operating regulations, for the necessary dumping of empirical memory and restoring the initial structure of the core once a year.
Working with the analyzer of the information field in a simplified version resembles watching television. You turn on the channel and feel everything that the owner felt, and of course you leave a corresponding mark on your information field ... It is relatively simple to view static universes ... the universes of already dead people or people diagnosed with brain death. Still alive create such a dynamic picture that it is impossible to make out anything ... It is like reading a book, the author of which constantly edits it in an arbitrary place and makes it thousands of times faster than you can read at least one word. Therefore, the universes of the living at our current technological level are not available. But science fiction writers are already dreaming not only of the possibility of reading the living, but also of changing the universes. Especially to raise this topic like fanatics dubious ideas. So the tales of the CIA satellites and foil hats from the 21st century returned with the same force and served the benefit of some enterprising "inventors" of personal firewalls of the information field, reliably protecting their owners from what is not yet available even on a theoretical level.
Reading the worlds of the dead has given a tremendous impetus to the development of many sciences. Of course, not all the dead were interesting personalities, but it was deciphering the images of Einstein, together with the modern understanding of some aspects, finally made it possible to reconcile the irreconcilable sections of physics. So, when the universe of the creator of the core of the AI ​​suddenly discovered, many predicted a speedy resolution of the Problem No. problem. But that did not happen. The worlds in it were more ordinary. After 100 years, only one managed to accomplish what many talented scientists were striving for. And to expect easy readability and interpretability would be naive. Inside the universes there are no pointers or reference catalogs ... These are worlds of images ... For example, the planet cemetery, many poets had such worlds strewn with monuments, verses on tombstones, spreading crowns of trees alive or dead ... and then - the planet is desert ... only sand ... First they even decided that this is a world of childhood nightmares or adventures, until they found skeletons all over the area of ​​the planet. Here it became clear that without psychologists in any way, a third was added to the two previously existing directions for solving the problem of AI nuclei - the analyst of the worlds of the inventor of the AI ​​core. For some particularly promising, special institutions were created, the main purpose of which was to determine the extent to which the world was involved in the invention of the AI ​​core or the threshold of self-development. For this kind of work required deep knowledge not only in the theory of AI, but also the psychology and biography of the inventor. And people with such qualifications usually preferred to join university groups to develop a new AI core without flaws, rather than delve into the worlds of children's dreams and look for a needle in the universe.
A door. Normal lock. Standard hinges. One single closed door in the palace with a thousand open ... One single building on the only planet of a single star in this sector of the universe ...
Source: https://habr.com/ru/post/417129/
All Articles