Quantum computer with artificial intelligence was first offered in Russia
Internet and PC protection with absolute resistance to cyber attacks
Projects – upgrade and millions of jobs
Scientific discovery – stochastic, self-developing system
Form of informational, intelligent life
System of international intercourse (SII)
The epoch of stochastics
Stochastics was developed in 1990s – during the first stage of artificial intelligence systems. It turned out that traditional technology cannot efficiently process the complex symbolic structures in environment of modern computers (based on Turing machine). As a result, logical conclusion – basis of intelligent systems – was not implemented. It was prevented by "combinatorial explosion" caused by brute-force method of its implementation. There was a new idea: bring the complex symbolic structures using stochastic (random) conversion to random numbers – combinations of set (certain) length, which ensure the desired (arbitrarily small) probability of collisions (at subsequent processing of both elementary symbolic structures and combinations thereof). These numbers were known as stochastic indexes. In fact, they were new unique names for symbolic structures of any complexity. The first property of stochastics (accident is uniqueness) manifested itself here. In the end, all structures get unique random names. Then, another thesis is applied – naming is understanding.
So, random indexes were provided with knowledge of original symbolic structures (in encrypted form). And here, absolutely new feature of stochastic systems – self-development and self-learning – was discovered. These systems could independently, without external influence, implement the logical and semantic links – such as "part-whole," "kind-type", "cause-effect", "condition-conclusion", etc. – by adding the random indices. For example, adding the stochastic indexes of words under mod.2 resulted in unique, random index of phrases – with arbitrarily small (set) probability of collisions. Adding the random indexes of phrases resulted in predicate index. Adding the predicate indexes under mod.2 may result in unique index of sentence or production rule, etc.
Based on that, they first obtained the self-developing intelligent environment, which could develop itself automatically, while forming the new knowledge. This fact implies the second and very powerful feature of stochastics – self-learning via automatic generation of meta-knowledge, which defines the logical connectivity of semantically related stochastic structures. As a result, we automatically get the implementation of non-brute-force inference method (despite the scope of original spatial search), which utilizes the logically related knowledge elements only. It enabled the solution of combinatorial explosion problem and creating the absolutely new generation of self-developing, self-learning, and efficient intelligent systems, which have no analogues in the world.
As noted above, stochastics or stochastic informational technology has the following features, absence of which prevents the understanding of meaning and elicitation of knowledge from unstructured texts:
- implementation of inference at large search spaces – using logically and semantically related text structures only – with exception of brute force method all over the search space, which ensures the exception of combinatorial explosion;
- implementation of self-development and self-learning of new knowledge, which defines the logically and semantically related text elements, and forming of new knowledge required to obtain the "world view" and semantic classifications in different areas of concern;
- BigData problem solving via automatic establishment of knowledge bases, which describe all possible features of concepts and logical relationships in the "world view" – in any possible situation, in search spaces no less than 1015;
- implementation of analytical and search functions at a set of original text information – using the obtained knowledge, in real-time mode, with maximum search space no less than 1020.
Only stochastics has the above features. They are unavailable for traditional information technology. Stochastics was developed in Russia (by Nasypny V.) and first published in monograph "Development of theory for building the open systems on the basis of AI information technology" (Moscow: Military Publishing, 1994. - 248p.).
In addition to the above features, stochastics has two more, which are relevant for current stage of IT development, namely:
- ensuring the guaranteed protection of created intelligent systems (operating in BigData mode) from information attacks – based on potential implementation of all search/analytical functions (in encrypted form) at a set of encrypted text arrays;
- full compatibility of stochastics with quantum technology, which enables the building of quantum computers (even now) with intelligence implementing the analytical and search functions at a set of search space no less than 1030.
Stochastics ensures the AI implementation on the basis of computational procedures of modern and advanced computers (RF patent No 2273879 dated May 28, 2002 "Method for synthesizing a self-learning system for extraction of knowledge from textual documents for use in search", international application No PCT/RU2002/000258 dated May 28, 2002, United States Patent Application 20050071150, Nasypny Vladimir Vladimirovich, March 31, 2005 "Method for synthesizing a self-learning system for extraction of knowledge from textual documents for use in search", China Patent Application ZL 02 8 29032.1, "Method for synthesizing a self-learning system for extraction of knowledge from textual documents for use in search", Priority 04.06.2008, etc.). Specific theoretical and technological solutions for creation of intelligent quantum computer are described in the following scientific papers: Nasypny V.V. "Development of theory for building the open systems on the basis of AI information technology" (Moscow: Military Publishing, 1994. - 248p.), Nasypny V.V. "Stochastics as a basis for transition to big data, knowledge industry, and nanotechnology" (Moscow: MPGU, 2011. - 24 p.), Nasypny V.V. "Stochastics. Perspective information technology. (Moscow: MPGU, 2012. - 106 p). In our opinion, colleagues from other countries beginning the research of application for quantum computer (to build the artificial intelligence) can rely on Russian experience in this field, which is of vital importance for development of the world science. In turn, we are ready to cooperate.
Big Data and Stochastics
The solution of the Big Data Problem is currently addressed as a prerequisite of the information technology evolution. The theoretical and practical aspects of this problem are prominently represented in paper . A few years before this the Stochastics, a stochastic and information technology, aimed for solution of the large bulk of data and knowledge processing problem has been developed in Russia.
This technology was developed from 1990 to 2000 [2-6]. A team of scientists created on its basis basic designs of an intelligent retrieval system , self-learning analytic system , integrated data protection in computers [2, 7, 11], continuous speech recognition and synthesis system for an unknown speaker’s voice [3, 8] etc. They all are proprietary designs both in Russia, and abroad. For the implementation of the intelligent retrieval system project the Stocona Research and Production Company in Russia and Stochasto in Norway have been established. This project was sponsored by multinational investment foundations. As a result, NearU, the first in the world intelligent retrieval system, based on the Russian stochastic information technology, was developed and installed in the USA (2005). This system provided a linguistic and semantic processing of large amounts of unstructured text information, data, and knowledge. The system employed areal time inductive logical inference over large amounts of data and knowledge, as well as knowledge acquisition from texts using self-learning algorithms .
Therefore, well before the Big Data Problem was understood in the West, particularly by IT leaders, such as Microsoft and Google, the Stochastics offered a solution in Russia.
A prototype of an Internet search engine in Russian and English languages for the implementation in Russia has also been created.
Several next projects are now ready for designing, specifically, speech recognition and understanding system, an intelligent retrieval system, a self-learning analytical system, proactive data protection in computers etc. Note that all those projects are in a varying degree solving the Big Data Problem, including such tasks, as accumulation, maintenance, and logic processing of large amounts of data and knowledge, knowledge acquisition from speech, images and unstructured text information. Self learning functions  as well as a deep analytic treatment of unstructured texts, data and knowledge is implemented.
These projects are unique and have no equivalents abroad. Their implementation in Russia would make a major contribution to the industry and entire economy modernization.
From the above said it follows that the Big Data Problem was raised and solved in Russia more than a decade before it was realized by leading western IT companies. The main point is Russia was the very country, where the problem of semantic interpretation and understanding of sensitive and text information as the basis for knowledge acquisition and deep analytical processing, as well for images recognition, was for the first time investigated and solved. Without such a solution any transition from the Big Data Concept to the knowledge industry would be impossible. This is the main advantage of the stochastics over existing western technologies.
Another most important merit of the stochastics includes a secure data and knowledge protection during transmission, storage and processing, as well as programs’ execution [2, 7, 11]. This is achieved due to the fact that the entire information is circulating, being processed in the computing environment in a stochastically transformed and protected form. And ultimately, stochastics, as we will show further, is compatible with nanotechnology, allowing implementing in the nearest future nation-wide information projects.
1. L. Chernyak Big Data: New Theory and Practice // Open Systems No. 10, 2011.
2. V.V. Nasypny Protected stochastic systems // Open Systems No. 3, 2004.
3. V.V. Nasypny Speech recognition and understanding based on stochastics in noise. Moscow: Prometheus, 2010. – 139 p.
4. V.V. Nasypny Development of the theory of open system integration based on the artificial intelligence information technology. M.: Voenizdat, 1994. – 248pp.
5. V.V. Nasypny, G.A. Nasypny Method for synthesizing a self-learning system for extraction of knowledge from textual documents for use in search. RF Patent No. 2273879, international application number РСT/RU02/00258, date of filing 28 May 2002.
6. V.V. Nasypny, G.A. Nasypny Method for synthesizing a self-learning system for knowledge acquisition for text-retrieval systems, patent application No. 2007120344/09 of 06.08.2007. Patent issue decision of 21.07.2008 obtained.
7. V.V. Nasypny. Absolute secure system // Open Systems No. 9, 2005.
8. V.V. Nasypny, G.A. Nasypny System for recognition, understanding of meaning, animation simulation and speech synthesis based on stochastic information technology. Moscow: Prometheus, 2008. – 76 p.
9. Artificial intelligence. Guide book. vol. 2. Models and methods. Under the editorship of D.A. Pospelov Moscow: Radio and communication, 1990. - 303 pp.
10. Halsall F. Data communications computer networks and osi. Addison-wesley publishing company, 1988. - 973 c.
11. V.V. Method for an integrated protection system of data distributed processing in computer networks and system for carrying out said method. RF Patent No. 2259639, international application number РСT/RU /00272, date of filing 28.10.2003г.
12. V.V. Nasypny, G.A. Nasypny Method for semantic linking of text and 3D graphics. – Moscow: Prometheus, 2007. – 27pp.