How to create artificial intelligence? History of the second. Algorithms for intelligent search and information storage
In a series of articles we tell about new approaches in Artificial Intelligence, modeling of personality and processing BIG Data that are not available for the majority of specialists in AI and the public. The value of this information is that it is all tested and most of the theoretical developments are implemented in the application projects.
All of you, without exception, the use of modern search systems that do well with search links by keywords. But can we consider these technologies, finding the information, are they able to answer your questions in different contexts if they are able to think, to imagine, based on the hundreds of thousands of sources in different languages?
Major search companies have almost unlimited resources they have money, they employ the most professional staff, thousands and even millions of servers service search queries.
So why until now you get as response a link, not an answer to your question?
In this article we will talk about the technologies and algorithms that will change your view on the search and storage of information.
In a previous article we introduced the basic terms, namely, intelligence, learning, imagination, memory, image, information flow, information field. Be sure to read it!: geektimes.ru/post/249042.
The article caused a certain resonance, and we received in three categories:
the
"is it Possible to write an algorithm that would be able to think, to reflect?"
"Do we have jurisdiction in AI?"
"What practical application we have developed?".
Today we tell about the algorithm of finding information on the Internet. In order to distinguish it from other algorithms we'll call it: "the core of v2.3". How the algorithm satisfies the criteria of intelligence for you to decide.
The basis of our algorithms formed the ASSUMPTION that the processes taking place during meditation, are electrical in nature. It can be assumed that when a person says something (thoughts in hearing), the various parts of the brain aktiviziruyutsya, and they aktiviziruyutsya not simultaneously, but sequentially. Different parts of the brain are responsible for different stages of the "maxims of thought". It can be assumed that the various "mental aspects" are dependent on each other.
Meditation (idea) is a narrative sequence of information units (including live images).
Considering that this assumption is true, we decided that the process of reflection can be simulated on the computer.
Consider this example:
"I'll come to you" — figuratively-information structure, it consists of three images:
the
-
the
- "the Image of the self" – "I"
- "Comprehensive or unifying image of what we want to achieve" – "to you".
"Way" – "Go", the
Thinking in different situations may have different connotation (hidden meaning):
For example, "I want to drink tea" in different contexts can have different meanings
the
-
the
- Who speaks this phrase, "the Woman says to the man", "woman to Woman", the
- At what time of day "evening", "day", "morning".
What in this situation "while walking", "office", "date", etc.
"the meaning" reflection – "it's the ability to read between the lines", from the information point of view, is the search for complex correlations (search informational and structural interdependencies).
We also came from the fact that any text — is a kind of logical sequence, which was somehow generated by the person during the reflection.
We decided to take a prepared text on the topic "How to easily quit Smoking", and search for this correlation, but found nothing. We hypothesized that learning occurs due to data redundancy, and took 500 articles on the topic "How to easily quit Smoking". Using various non-linear algorithms, we found complex correlations.
In particular, we found the points at which "the line of application of reflection" may be sent along different paths. These points we call "turning points" or "Motivators". We mentioned "Motivators" in the previous article, it was used the more popular term from psychology – the anchor.
"the Motivator" is a structural element of the chain, with "resistance" and allowing for the subsequent change of direction signal (thoughts) under the action of additional factors (motives). So the motivator can change the path "reflections", if in the process of reflection will be "additional capacity".
Another important property of "pivot points"/anchors/motivators – use them as "entry points" or "moment of thought". In this way of thinking is consistent And if you start thinking with a different concept (NOT motivator), then the first thing that happens is the alignment to the motivator. In life this manifests itself in the following way. If you are asked to reflect on something unexpected for you, then you will most likely begin their argument "far out". By meditating, you must reach some point(motivator), and then continue his meditation in the "given" direction.
In order to process and store such information structures, it was necessary to create a special memory model, we have designed and called "factorial".
"memory Factor" is a structure for forming an information field from a flat infopotok, as well as the definition and management of motivators. In other words, the factor model memory is memory that uses a semantic image representation as a set of characteristics, factors, and combinations of the ordered data – the facts.
We remind you that from our point of view intelligence – the ability to learn and imagine.
Let's think whether the described model of the nucleus these qualities?
self-study, learning goals, criteria for training in the context of the model of AI.
Learning is the transformation of information flows in the information field.
The model has the main motive to solve the problem, cope with the difficulty coming from the outside.
Its task is to recognize patterns, which come in the form of a query, to create a complete image, using redundant information in the Internet.
So as to request the memory model clean, it needed to "learn the language" (the language is not explicitly defined), to understand the meaning (to find complex correlations and motivators), to form a redundant information field in order to be able to think (to create an information chain).
Main technical the criterion of "filling memory"/"creation of the information field/create integral image" is the estimate of the entropy in the information field. Entropy is a criterion of adequacy that the image is assembled, and at the same time the main criterion of lack of information.
Learning isn't possible without feedback
Feedback in the model v 2.3 is as follows – user can adjust "the starting point of reasoning" and choose the direction of formation of the response(reflections) within the formed information field. The model demonstrates the result of reflection to the user and waits for corrective specifying word (motivator).
Imagination in the context of the model AI
Imagination we have defined as the ability to finish the unfinished image on the one hand and the ability to simulate the situation on arbitrary conditions, on the other hand.
The algorithm is capable of activating different motivators, different chain of thought.
Examples of model kernel v 2.3
1) Request: "the Impact of violent computer games on children".
Context clarification: Without specifying.
Reflection models:
"in the Spring of 1999 bill Clinton in a televised address to America said: "We have to think twice when it comes to "first-person shooters"".
Hot on the trail in the April issue of the journal "Personality and social psychology" for the year 2000 published an article "video games and aggressive thoughts, feelings, and behavior in the laboratory and in life".
The authors, Professor of psychology Craig Anderson and Karen dill, argued: kids see more violence in the world today.
The article claimed that there is a connection between violent video games and aggressive behavior of adolescents that can lead to crimes and murders.
Anderson said: "Although in the study of behavioral characteristics there are many difficulties, one simple and clear thing you need to know everything: computer games increase violence.
Our study showed that the influence on the aggressive behaviour in violent computer games are even worse than violent TV and movies".
What if a step to move away from our vegan conventionally difficult situation and think about someone playing in these death kids will grow.
Many experts disagreed with this opinion, believing that the conclusions of Anderson and Dyl opinion rather than proven scientific fact.
26 April 2002, Robert Steinhauser killed 17 and wounded 7 people in the gymnasium Gutenberg, Erfurt, Germany.
The authoritative newspaper the Frankfurter Zeitung Allgemeine came out with an article "software for the massacre", in which he wrote: "the killer was trained using computer games".
The document explicitly prohibits "the dissemination of all types of products that promote violence, aggression and antisocial behavior".
Responsible for the development of actions was assigned to the Office of youth Affairs of the regional administration.
The aim of all these activities has been the struggle against youth aggression and violence.
These events caused a reaction in the neighboring areas of Ukraine, where urgently Belgorod adopted a similar resolution on the fight against violence, and representatives of law enforcement agencies began to seize the "violent computer games" in the clubs.
Hard computer games rather discharge the psychic energy accumulated during aggressive self-assertion of teenagers in everyday life.
With this increased activity and aggressiveness of a teenager in relation to the outside world, she gives him support to overcome his own fear.
Frankly, I myself once upon a time very fond of all sorts of shooting, I never noticed aggression."
The request can specify a word or part of a word. For example, "young", "games", "cruel" etc., after further reflection will change.
The motivators of the information field:
year---play---act---
resolution---against---propaganda---
brain---which---answer---
frontal---shares---person---
computer---games---help---
computer---games---parents---
2) Request: "Problems of artificial intelligence".
Context clarification: "problems".
Reflection models:
to Identify the structure and understand the principles of organization of intelligence — it means to reveal the real reason of the fundamental problems, to show her the deep specificity.
In other words, it is necessary to examine the historical, scientific, technological and epistemological aspects of the problem of artificial intelligence with the direct participation of the dialectical materialist philosophy.
Philosophy, chose as the main question of their scientific researches the problem of the relationship of existence and consciousness, should be more closely interested in the issues related to the modeling, the reproduction of intelligence.
Artificial intelligence is directly related to the basic philosophical question: we are talking about artificial reproduction, the simulation of human intelligence.
In addition, the first work on artificial intelligence has shown that not only the solution scope task dealt with comparative research, but the problem of thinking in General.
Different types of animals we find the nervous system of different complexity, and if we gradually move along a ladder of complexity in the evolution of understanding what mechanisms are responsible for a particular leap in intelligence to animals, then perhaps this will allow us to solve the problem of understanding human intelligence.
The motivators of the information field:
Also---solution---problem---
---study---study---
human---activities---activities---
Also---problem---solution---
region---artificial---intelligence.---
problems---artificial---intelligence---intelligence---
region---artificial---region---artificial---intelligence.---
3) Query: "how to find your meaning of life"
Context clarification: "life"
Reflection models:
"In his theory, Frankl focuses on the people who are dependent on alcohol and drugs, which see not only the meaning of life, but generally the real life.
What is the meaning of life of a specific person?
If there is one the meaning of life for all people on earth, or it is unique for each?
And if you did not get to enjoy, then you have lived a pointless life.
In this regard, on any vital interval it is important to remember that our path has not only a beginning and an end.
Everything is done in your life is important because of the person remember long after his death.
The meaning of human life is to develop from the primitive mind of the microorganisms to Superintelligent beings of the distant future.
But the most interesting question whose answer we will never know: the meaning of life in General.
Happy people such questions are not asked, they just live and enjoy life itself, therefore these questions come to mind when man is unhappy.
The meaning of life is the same for all living beings, from microbes to humans: to survive and to leave offspring.
And even cooler — life is just a very complex chemical reaction mostly of carbon atoms occurring in nature, irrespective of you.
The purpose of life is life itself, realizing its divine essence!
Contrary to popular belief, the lives of people ruining not low self-esteem and inflated, derive from it the ambition, in the absence of implementation of which the person begins to feel pain.
The meaning of life in all living things, one: the DEVELOPMENT of discernment of good and evil, benefit from harm, the true knowledge from errors.
Since the purpose of all in this life of their own.
It is necessary to listen to your feelings, not your thoughts and understand what I want to do with my life.
For this we need to get away from the everyday hustle and bustle and to stop looking for the meaning of life.
That is incorporated and in which direction in life to be realized.
Having found his calling, will begin to enjoy life and every moment.
But life brought to the original."
The motivators of the information field:
yourself---this---life.---
find---yourself---yourself---
the meaning of---life---life---
life---everyone---life---
algorithm Performance
Our research began more than 5 years ago and took place on a dedicated cluster. After certain laws were found, it was necessary to optimize the algorithms. Several years were required in order for meditation to work on normal server hardware. Kernel version 2.3 even able to reflect on your tablet with the Intel Atom processor.
Model core 2.3 is a single cell of memory (algorithm "a single-Celled memory"), one information structure that has sufficient redundancy to ensure that the framework can determine a way, and under different points of view to present. The prototype model of the nucleus 2.3 we have uploaded for testing.
Opportunities and the development of the DES algorithm
If we use memory from multiple cells, we can build a more complex correlation in such models, there are transitions from one memory location to another. Therefore, such algorithms are called deductive or detektorami or DES. Unfortunately, these algorithms are little known to the public, as usually such studies are closed.
As infopotok can be any stream of data: text, video, telemetry, etc. of the stochastic Algorithm is able to find correlations and build the sorted data in chronological order, and without it. The use of text data allows you to simply get's an information flow and to demonstrate the results of the algorithm.
If you want to experience the capabilities of kernel v2.3, the ask us link through dialogues. We do not spread the link in public access, because the model kernel is installed on a weak server, and would like to give direct advertising.
In order to implement such technology EN masse, it is not enough to possess them, you must also change the informational paradigm, and create new approaches for processing and storage of information.
In the next article we will discuss the next generation technologies of the kernel is already multi-celled memory model, allowing to simulate some aspects of personality, dialogue in the framework of the personality model of the user. Similar technology can be used to create personalized web services, applications, helpers.
Learn and imagine...
Комментарии
Отправить комментарий