The paper addresses the issue of forming an intelligent digital infrastructure for managing the country’s economy in special conditions: global disasters, catastrophes and emergencies. The experience of creating large information systems for managing the country’s economy in a special period is evaluated. The convergent approach for creating the required system for management support is proposed. It is shown that under these conditions, high management efficiency can be achieved on the basis of special support for self-organization processes and anti-collapse self-adjusting integration of information system segments and intelligent services that are adaptable to conditions of a known, predictable and unknown nature. At the same time, the integration of network infrastructures involves the distributed processing and storage of data based on interaction and integration of various network environments, which allows to achieve previously unattainable reliability, stability and recoverability of economic management.
The article by Gary Marcus Deep Learning: A Critical Appraisal  dwells on the questions about the current achievements of in-depth training and artificial intelligence (AI). The overall tone of the work is pessimistic and tends to rethink the results, even if they are intermediate. Markus makes forecasts and writes about the possible consequences of another hype around AI technologies . The article gives a vision on which direction to move while developing AI systems.
IT is the fastest growing sector of the economy, where information is the product: description of technologies, software and services, digital multimedia material, e-books, patents, etc. Payment systems also represent a kind of information. If previously the period of technological structures change constituted hundreds of years, now it happens during one person’s life. Gartner company conducted a survey and published the list of priority directions of IT development in 2017: artificial Intelligence and deep learning; intelligent applications; intellectual things; virtual and augmented reality; digital counterparts (dynamic program models of real objects and systems); blockchain (cryptocurrency) and distributed expenses book; conversational systems; network applications and architecture; digital platforms; adaptive security architecture.
The course of artificial intelligence development in the XXI century in the context of formation of new technological modes and post-nonclassical management paradigm can be forecasted based on analysis of its many-sided and thorny historical retrospective. This way, of course, is not assured from unexpected traps. However, they can be avoided by finding new solutions in unfamiliar spaces for modeling, based on other approaches to solving complex problems and semantic interpretation of data by synthesizing for these innovations still unknown materials for computer memory and processors.
The possibility of artificial intelligence was originally associated (von Neumann) with the problem of overcoming some hypothetical complexity threshold. Currently, the formation of a modern complexity paradigm in the context of philosophical ideas of E. Morin , G. Deleuze and F. Guattari , second-order cybernetics of Heinz von Foerster, autopoiesis of F. Varela and F. Maturana, cibersemiotics and recursive logic of “laws of form” by J. Spencer Brown causes the necessity of constructive introducing the concept of complexity observer as a self-organizing ensemble of cognitive agents; the artificial intelligence and artificial consciousness will probably become an emergent product of their interaction.
The paradox of organizational structure of modern companies is that people want to do something and interfere themselves with one other.