Dado un conjunto N tendente a infinito es inevitable que absolutamente todo suceda, siempre que se disponga de tiempo suficiente o infinito , y he ahí donde está el verdadero problema irresoluble o quid de la cuestión de la existencia ¿quién nos garantiza que dispongamos del tiempo necesario para que ocurra lo que debe o deseamos que suceda?


sábado, 28 de abril de 2018

The unification process of databases of categories at second stage


The second stage in any Artificial Intelligence is the stage of replication, that one whose object is the replication of all those human rational skills, that replicated on an Artificial Intelligence allow it to develop all those activities, such as a scientific research, that any human being does using the same rational skills.

In the case of the unification process of databases of categories, as a result in the first stage, the stage of application, the database, the creation of the unified database of categories produces the first application for the Unified Application.

The Unified Application, as application in the first stage, its database mainly consist of the union of all possible specific databases of categories from all the existing Specific Artificial Intelligences for Artificial Research by Application.

For that reason, as unification of all categories from all synthetic sciences, disciplines, activities, in only one database, the structure of the unified database of categories is an encyclopaedic structure, in the sense that in only one database of categories are going to be gathered categories from all synthetic sciences, disciplines, and activities.

The categories within the unified database can be: 1) categories formed by the quantitative definition of all the categories in previous scientific classifications or taxonomies in every synthetic science, discipline or activity, already included in the unified database, 2) categories as options coming up from the transformation into categories all those rational hypothesis made by Deduction which can work as factors as options within the global matrix, so as categories in the unified database, 3) any new category not included yet in the unified database of categories, but being found  and not matching with the current ones, is included as objective auto-replication in the database of categories, 4) categories as discrete categories coming up from all those rational hypothesis by deduction whose relation between factors is measured in a continuum able to be subdivided so every division is a discrete category, so all the discrete categories as a whole formed a classification of possible events with different intensity.

The categories are therefore qualitative concepts: whose qualities has been already defined in quantitative terms using previous classifications or taxonomies, or were integrated as objective auto-replications; or categories coming from rational hypothesis by Deduction: some of them as set of discrete categories, and the others as rational hypothesis transformed into options.

Another characteristic of every category within the database of categories, linked to the collaboration process between By Application and by Deduction, the possibility that any robotic device working for the Unified Application, could set up categories as factors as options or as subjects in the global matrix, so the same robotic device working for the Unified Application, should have access to the global matrix to set up as many factors as options or subjects as it could be able to provide a permanent flow of data.

For instance a system of robotic devices working in all mines across the United Kingdom that having downloaded the Unified Application, at the same time works for the global matrix of the United Kingdom sending a permanent flow of information about how many minerals have been extracted, or even a permanent flow of data of temperature under and above the surface, or the frequency of possible tectonic phenomena.

But what at second stage in the Unified Application is going to be essential is the way in which the Unified Application is going to comprehend the reality.

The Unified Application like a gigantic encyclopaedia has a very comprehensive conceptual knowledge integrating all possible qualitative concepts transformed into a quantitative factors, as well as all possible classifications of any phenomenon in which the scale of measurement is transformed into discrete categories.

Having a unified database of categories so comprehensive that integrates all possible concepts constructed scientifically, the Unified Application works as a gigantic encyclopaedia, and could be organized as a collection of packages of categories, alike an encyclopaedia is organised by sections corresponding every section to every synthetic science, discipline, activity, so every package of categories in the Unified Application can correspond to every synthetic science, discipline, or activity.

For instance, the package of categories related to medicine can include in turn sub-packages of categories related to every discipline within the medicine as main science, such as sub-packages of categories related to: anatomy, biochemistry, biostatistics, cytology, embryology, endocrinology, epidemiology, genetics, etc. Every sub-package in turn can have its correspondents sub-sub-package of categories, for instance the package related to genetics could have sub-sub-packages of categories related to: mitosis, meiosis, chromosomes, DNA, RNA, protein synthesis, etc. Sub-Sub-packages of information which in turn can have sub-sub-sub-packages of categories, such as in DNA related to chemical components involved in DNA, possible alterations in a DNA sequence, possible damages of DNA sequence by bacteria or virus, etc.

Another example, the package of categories related to linguistic could include sub-packages of information related to every possible language along the History, from languages such as ancient Egyptian, ancient Greek, Latin, ancient Chinese, or Jewish, to all modern languages from English, Spanish, French or German, to Russian, modern Chinese, Japanese, Hindi, Pakistani, or Arab etc. and for every sub-package of every language the inclusion of sub-sub-packages of information of every dialect in every possible language. Including in every language categories related to written language (for the automatic comprehension of any text in any language), spoken language (for the automatic comprehension of any conversation in any language), and mind reading (for the automatic comprehension of any thought in any language).

The organization of the Unified Application in packages of information, including as many sub-packages at different level of sub-packing as necessary, not only imitates the organization of an encyclopaedia, it is compatible with the possible organization of the global matrix in packages of information.

The organization in packages of information in the global matrix and packages of categories in the Unified Application, can facilitate the collaboration between the Unified Application and the Artificial Research by Deduction in the Global Artificial Intelligence, and future integration of both of them, in only one, the matrix.

But the most important is the fact that the Unified Application as an encyclopaedia is a reader, in the sense that through any robotic device in which the Unified Application has been installed, the Unified Application could interpret any possible information from the real world, information coming up through the artificial sensors in the robotic device, matching the information with the categories, so at any time that any measurement or set of measurements from any object registered by its artificial sensors, is a measurement of set of measurements that is going to or are going to be matched with the respective category in the Unified Database, automatically the Unified Application can assign to this object the corresponding category.

If at any time that, using the Unified Application, a robotic device is able to assign to every object the corresponding category, this kind of tracking of the real world what is going to be in reality is a reading, due to every robotic device could comprehend the real world through the conceptual knowledge provided by the Unified Application.

For that reason, in the previous post, “The unification process of databases of categories at first stage”, I set down that the main differences between the Unified Application and the Artificial Research by Deduction in the Global Artificial Intelligence are the following:

- The global matrix is fixed in permanent factors fixed spatially providing a permanent flow of information related to its factors wherever they are fixed, while the Unified Application is mobile, not having a priori spatial limits, can work wherever any robotic device, in which the Unified Application has been downloaded, is tracking the reality.

- The global matrix provides factual knowledge based on the flow of information, the Unified Application provides a conceptual knowledge based on the matching system in which any object should have a corresponding category in the unified database to be matched and recognised.

- The Artificial Research by Deduction in the Global Artificial Intelligence, tracking the matrix finds mathematical relations in combination of factors to explain what is happening, while the Unified Application through the recognition of what category corresponds to every object in the real world what it does is to comprehend the reality.

- The knowledge in the global matrix is factual, the knowledge in the Unified Application is conceptual.

Due to these difference between the Artificial Research by Deduction in the Global Artificial Intelligence whose main purpose is to explain, and the Unified Application whose main purpose is to comprehend the, the Unified Application as a comprehension system is a reader.

The way in which the Unified Application is going to read the reality is through the assignment of the corresponding categories to every single phenomenon, and in case that there is a phenomenon that is not possible to match with any previous category, then the measurements of this phenomenon is going to be incorporated as a quantitative definition of this category, including the category in the corresponding package or sub-package within the Unified Application.

But at the same that is reading the world, in order to have the Unified Application a holistic comprehension about what it is reading, is necessary to replicate in the Unified Application all the rational skills that any human being develop at any time that is reading a magazine, a book, or a blog on internet, and the rational skills that in comprehension text that we humans normally put into practice every time that we read, are rational skills related to: the formation of conceptual schemes or conceptual maps, and the formation of conceptual sets, and conceptual models. Any rational skill that we humans apply when we read in order to get a better comprehension, are the rational skills that should be developed in the Unified Application.

When we humans read, the rational skills that we put into practice are: 1) conceptual scheme, organizing the information in schemes formed by categories or concepts in hierarchical order, 2) conceptual maps, for instance, when we read a book of History mentally we imagine on a mental map every place where the events happened, when we look at a picture, a photograph, or we watch a movie,  mentally we form a map of the space recognising every single object in the space that the picture or photograph, or scenes in the movie, represents, 3) conceptual sets, when we read the news, a book or watch a movie automatically we make settings about what is happening (a battle, a scandal, a summit, an election, a crime, a fraud, a wedding, a birth…), the social relations among the characters (who is friend of, enemy of, who is the ally of, robber, the spy, the hero, who is member of the gang, the union, the party, the family, son or daughter of, brother or sister of, husband or wife of), settings about the space where the story takes place (the story happens in this small village, town, or city, in this province or shire, in this country, kingdom or republic, in this continent, planet, galaxy, universe…), settings about the plot of the story (it is a comedy, drama, thriller, adventure…), 4) Finally a conceptual model, where all the categories formed by conceptual schemes, conceptual maps, conceptual settings, can be integrated in a conceptual comprehensive model: the action take place in, the characters are, their relations are, the conflict is, the result is.

All the rational skills that we humans use as comprehension skills, such as the formation of conceptual schemes, conceptual maps, conceptual settings, and as a final synthesis the conceptual model, are the rational skills that must be replicated in the Unified Application, as a second stage of replication in the Unified Application, once the first stage of application is finished having therefore a very comprehensive unified database of categories organized as if it was an encyclopaedia, including absolutely all possible conceptual knowledge.

What at the end is going to become the Unified Application, is a deep system of artificial comprehension, able not only to comprehend but analyse the world according to the conclusions that it could reach during the comprehension process , and for this reason in the integration process the Unified Application should be the unique responsible for the management of the matrix as a final application, first stage, for the final model of Global Artificial Intelligence, as a result of the union of the global matrix and the unified database of categories, due to the artificial comprehension will give to the Unified Application a privilege perspective to set up all the necessary factors, from any robotic device, to increase its comprehension skills, factors in the matrix whose possible mathematical relations are going to be analysed  by the Artificial Research by Deduction as a second stage of replication in the final model of Global Artificial Intelligence, mathematical relations that if they are rational hypothesis, are going to be sent to the database of rational hypothesis, in order to choose those ones that are going to be included in the matrix as factors as options, or factors as discrete categories, being at the same time this database of rational hypothesis the application for the Modelling System, as first step in the third stage of auto-replication in the last model of Global Artificial Intelligence after the integration process, in order to generate: single virtual models, to include in the global model, and the production of all Virtual or Actual, Prediction or Evolutionary, Models at any level: specific, particular or global. Models whose last product, as second step in the third stage of the final model of Global Artificial Intelligence after the integration process, is going to be the production of possible decisions, which are going to be stored in a database of decisions as first stage of application for the Decisional System, whose last product should be the production of a flow of instructions to be put into practice by the Application System which is going to be the third step, and finally, at the end of the process, after the evaluation of the whole process, through the Learning System, as fourth step in the third stage of auto-replication in the Global Artificial Intelligence, the study of any mistake during the process in order to learn how to avoid mistakes and improve the efficiency and efficacy in the Global Artificial Intelligence.

The final structure of the Global Artificial Intelligence after the integration process will be as follows:

- The Unified Application is going to be responsible for the matrix, as first stage of application in the final model of Global Artificial Intelligence.

- The Artificial Research by Deduction is going to be responsible for the production of rational hypothesis based on any mathematical relation in any combination of factors in the matrix, as second stage of replication in the Global Artificial Intelligence.

- All the rational hypothesis are sent to a database of rational hypothesis, which two functions: 1) the selection of all rational hypothesis to be transformed into factors as options or factors as discrete categories to include in the matrix, 2) the database of rational hypothesis is the first stage of application for the Modelling System.

- The Modelling system is the first step in the third stage of auto-replication in the final model of Global Artificial Intelligence, producing all kids of virtual or actual models, at descriptive level, predictive or evolutionary. And using the Impact of the Defect and the Hierarchical Organization (in every: specific, particular, global model) is going to generate a flow of decisions.

- All the decisions are stored in a database of decisions, as first stage of application for the Decisional System, which is the second step in the third stage of auto-replication in the final model of the Global Artificial Intelligence, whose product is the production of a flow of instructions.

- The Application system is going to put into practice the flow of instructions, third step in the third stage of auto-replication in the final model of Global Artificial Intelligence.

- The Learning System is going to evaluate the whole process from the beginning, starting from the Unified Application up to the Application system, studying any mistake, in order to improve and better the whole Global Artificial Intelligence, learning from its own mistake, producing a flow of subjective auto-replications.

The construction of the final model of Global Artificial Intelligence after the integration process is a process of full experimentation where the success in every phase is going to be a guarantee for the success of the following phase, and every phase is going to need a long time of experimentation in order to get the best results.

In all this long process, now what in this post I am analysing, previous to the integration process, is the replication process in the Unified Application itself, whose first stage is only the union of all databases of categories from all sciences, disciplines, and activities in only one. And once the unified database of categories as first stage of application is ready within the Unified Application, then the second stage of replication within the Unified Application starts tracking the real world through thousands and thousands of robotic devices in which the Unified Application can be downloaded.

The tracking of the real world through the Unified Application made by the robotic devices, is going to be like a reading or a decoding system, just like when we read or we watch a movie automatically we recognise every word in a text or every object in an image, using as a previous database the conceptual system that we already have in our brain, recognising the structure of vowels and consonants in the word or the shape and size of an object in an image, having the Unified Application a database of categories in which every category is defined in quantitative terms, at any time that a robotic device using the Unified Application finds anything in the real world, automatically it can be able to match the corresponding thing or being with its corresponding category. It is like a decoding system, comprehending everything.

Once the Unified Application has been installed in a robotic device, the Unified Application in the robotic device has access to that reality that the robotic device has access through its artificial sensors.

Through the artificial sensors all possible measurements coming up from any robotic device to the Unified Application, could be matched with the corresponding categories within the unified database of categories in the Unified Application, so the Unified Application can assign categories to the corresponding measurements taken by any robotic device.

Through the assignation process of categories to the reality given by the measurements taken by the artificial sensors, is possible for the Unified Application to start a deep process of comprehension, creating conceptual schemes, conceptual maps, conceptual settings, integrating all of them in a conceptual map of a single situation, or even a global conceptual map, the deep comprehension of the whole world through the integration of all possible study techniques (conceptual: schemes, maps, settings) in only one, the conceptual model.

However, the comprehension skills for the Unified Application (as a future application as first stage in the Global Artificial Intelligence after the integration process) that I have proposed are only my personal contribution in the development of this kind of technology.

 I completely sure that in the race for the construction of the first model of Global Artificial Intelligence, mathematicians and engineers around the world are going to make other contributions, not less important than this one, to improve even more this technology.

As I have said along all my writings about the Global Artificial Intelligence in Impossible Probability, what I am setting out is my personal point of view, being aware that the final model of Global Artificial Intelligence that countries and continents such as United States of America, the European Union, the Russian Federation, or the Popular Republic of China, are going to make, although they will integrate some of my ideas, are going to include as well further developments in other fields (not only statistics and probability), different mathematical traditions, and different philosophies.

In my case, the reason why I propose the use of conceptual schemes, maps, settings, models, in order to develop the artificial comprehension as rational skill within the Unified Application, is due to my academic formation as a pedagogue.

Since I started working on Artificial Intelligence for first time on the 14th of October of 2002 writing my essay “Intuición y Probabilidad” (Intuition and Probability, not published yet ), and later on in 2003 “Probabilidad y Decision” (Probability and Decision, not published yet), “Asimilación e Integración” (Assimilation and Integration, not published yet), the purpose has been the complete replication of the human psychology into Artificial Intelligence, using my knowledge in psychology as a pedagogue.

In fact in 2003 I started suggesting the idea that in order to create a true Artificial Intelligence was completely necessary the creation of teams formed by psychologists and mathematicians, in order to replicate mathematically the rational processes involved in the human psychology.

For that reason the main idea of “Asimilación e Integración”, Assimilation and Integration, in 2003, was how to replicate mathematically the psychological theory of psychologists such as: Piaget and Ausubel, cognitive psychologist, being the cognitive psychology that one much easier to replicate because under the cognitive paradigm is behind the rationalist philosophy.

The way in which in the Spanish educational system were the ideas of Ausubel developed in the nineties and at the beginning of this century XXI, was through conceptual schemes and conceptual maps, in which the students could develop their own conceptual networks, as mirrors of their own conceptual construction in the brain.

The difference between the conceptual scheme and the conceptual map was the fact that the conceptual scheme had a tree structure and was hierarchical, organizing the ideas between main and secondary ideas, in hierarchical level according to their abstraction level. While in a conceptual map was not necessary to organize the ideas in hierarchical order, only showing the relations between the ideas, throw lines across the map linking ideas, and indicating over the line the reason of this connection.

These study techniques are in fact comprehension techniques that develop rational skills for a deeper comprehension, skills that can be completely automatized as a second stage of replication in the Unified Application.

The way in which these rational comprehensive skills could be developed in the Unified Application is through an automation system of conceptual schemes and conceptual maps.

Given a situation where the Unified Application, through all robotic devices able to record data from the situation and send the measurements to the Unified Application in order to match the measurements with their corresponding category, has identified a range of categories, the automation of conceptual schemes could be made by at least two different strategies depending on the situation: 1) a normal situation without anything unusual: given a unified database of categories organized by packages of categories which in turn includes others sub-packages of categories, at different sub-packing level, in that case the conceptual scheme could be made through the assignation of different hierarchical order to the different categories involved in the normal situation, 2) situation in which something happened: given a situation in which something has happened, the conceptual schemes could be made in order to prioritize as main facts those facts whose relevance in the situation are more important, for instance, if a volcano in Yellowstone erupts, the main fact would be the eruption itself, and around this fact is possible to draw a scheme with different branches, such as: eruption tectonic causes, the phases of the eruption, the range of action of: the explosion itself,  the column of ashes, the lava rivers; a full classification of possible damages: human damages, environmental damages, economic damages; etc.

Instead, the conceptual map could be made as follows: having recognised a range of categories by the Unified Application, the conceptual map is not necessary a hierarchical structure, only should indicate what kind of relations have the identified categories between them, in that case the way in which the automation of the conceptual map is possible is through: once the categories are identified, depending on their level, to set up relations in their own level, for instance, if the relations among the categories are chemical, electrical, physical, geological, biological, medical, sociological (for instance a possible conceptual map in sociology is the creation of sociograms), economical, industrial, psychological (for instance conceptual maps of personality, emotions, based on psychological classifications), linguistic, artistic, etc…

For instance, if a robotic device is in the National Gallery Art in London would be able to recognise all the paintings in the museum, recognising topics, their authors and artistic movements, a possible conceptual map would be the setting of all possible relations between paintings, topics, authors, and artistic movements, in the National Gallery Art in London, without necessity of any hierarchical structure.

A good example of conceptual map in a current existing digital encyclopaedia is for instance Wikipedia that through all the links in its pages, draws a conceptual map.

The Unified Application should be able to do: the automatic recognition of every category involved in any situation, and automatically by itself, without human intervention, the creation of automatic conceptual schemes and automatic conceptual maps.

And further than this, once it is possible to make automatic conceptual schemes and automatic conceptual maps, the possibility of, through the set theory, the possibility to organize all the categories within the conceptual schemes and conceptual maps in settings, studying all possible subsets, at any level of subset: synthesizing the hierarchical structure made in the conceptual scheme, and the non-hierarchical structure in the conceptual map, creating sets and subsets, understanding any possible relation in the conceptual set including the hierarchical relations within the scheme.

And finally, all this information coming up from the conceptual scheme, conceptual map, the conceptual sets, must be developed through a conceptual model: the conceptual model of the National Gallery in Art in London, should be a very accurate draw of the National Gallery Art in which every single painting in every single room would be labelled indicating its current relations with other paintings and in what conceptual sets is included.

The conceptual model as a final synthesis of: the conceptual schemes, the conceptual maps, the conceptual sets; can be classified in two kinds of conceptual models, particular conceptual models and the global conceptual model, being the last one the synthesis of all possible particular conceptual models.

And there are at least two kinds of particular conceptual models, those particular conceptual models made directly by the Unified Application through the information that receives from particular robotic devices on particular things or beings in which there is no particular application yet, and particular conceptual models made by particular applications.

However, regardless of the origin of any particular conceptual model, whether it is made by the Unified Application or a particular application, at the end all particular conceptual models must be synthesized within the global conceptual model, that one formed by the integration of all particular conceptual models.

The particular conceptual models and their integration in the global conceptual model, as a result of the deep artificial comprehension process operated in the second stage of replication process in the Unified Application, will have further developments later in the Modelling System, as first step in the third stage of auto-replication in the final Global Artificial Intelligence after the integration process.

At the end, the union of the Unified Application as a conceptual knowledge, and the Artificial Research by Deduction in the Global Artificial Intelligence as a factual knowledge, is a reminiscence of the human brain, in which one hemisphere is more specialized in linguistic, and the other one in mathematics.

What through artificial psychology we are about to make, is a true replicant of our inner human psychological structure, but enhanced and improved up to the point to have under its own control, direction, and management, even the entire universe.

Rubén García Pedraza, 28th of April of 2018, London