Limits and Risks of Digital Transformation

Currently, the process of digital transformation is actively going on in the economy, science, education, and society as a whole. This process has a number of restrictions and risks we consider. The mathematical theory of complexity reveals a large class of the restrictions. The exact solution of a number of simple-looking problems with a small amount of input data requires resources many times greater than the capabilities of all available computers.

On the "border" between natural and artificial intelligence lies the "cognitive barrier". This, as a rule, makes it impossible to use the results of a number of artificial intelligence systems to adjust our strategies. We and computers "think" differently. They have to be considered as "black boxes". It is very likely that the tester of artificial intelligence systems will become one of the mass professions in the close future.

We give examples to show that the "translation" from "continuous" to "discrete" language can lead to qualitatively different behavior of mathematical models. In a number of problems associated with a computational experiment this can be quite significant.

Great risks arise when passing to the "fast world", approaching the "Lem's barrier". It happens when artificial intelligence systems are assigned strategically important tasks that they must solve at a speed inaccessible to humans.

The analysis shows that managing the risks of digital transformation and its limitations requires the attention of the scientific and expert community, as well as active participants in this process.

1. Schwab K. Shaping the fourth industrial revolution. Portfolio Penguin, 2018.

2. Decree of the Government of the Russian Federation of July 28, 2017 #1632-r. Programma "Cifrovaja ehkonomika Rossijskoj Federacii".

3. Vvedenie v kriptografiju.Ed. V.V.Jashhenko. MCNMO, 2000.

4. Komarov M. M. Fizicheskoe proektirovanie submikronnykh SBIS. Problemy, zadachi, algoritmy. Budushhee prikladnoj matematiki: Lekcii dlja molodykh issledovatelej. Ot idej k tekhnologijam. KomKniga, 2008.

5. Vladimirov V.A., Vorobev Yu.L. et al. Upravlenie riskom. Risk, ustojchivoe razvitie, sinergetika. Nauka, 2000.

6. Turing A.M. (1950) Computing machinery and intelligence. Mind LIX(236), pp. 433-460.

7. Lee K. F. AI Superpowers: China, Silicon Valley, and the New world order. - Houghton Mifflin Harcourt, Boston New York, 2018.

8. Knyazyatov S.A., Malinetskiy G.G. The solution of the problem of bluff detection in the game "I-doubt-it" based on reinforcement learning// KIAM Preprint. 2018, #170. [In Russian]

9. Malinetskiy G.G., Potapov A.B., Podlazov A.V. Nelinejnaja dinamika: Podkhody, rezul'taty, nadezhdy. URSS, 2016.

10. Tikhonov A.N., Arsenin V.Ya. Metody reshenija nekorrektnykh zadach. Nauka, 1974.

11. Knuth D.E. The art of computer programming: Vol.2. Seminumerical Algorithms. Addison-Wesley, 1969.

12. Samarskiy A.A., Popov Yu.P. Raznostnye skhemy gazovoj dinamiki. Nauka, 1975.

13. Rezhimy s obostreniem. Ehvoljucija idei: Zakony koehvoljucii slozhnykh struktur.Ed. G.G.Malinetsiy. Nauka, 1998.

14. Platonov A.K. Problemy i perspektivy robototekhniki/ Budushhee prikladnoj matematiki. Lekcii dlja molodykh issledovatelej. URSS, 2005. P.315 342.

15. Lem S. Weapon systems of the twenty first century or the upside-down evolution. 1983.

16. Malinetskiy G.G. The risks of digital reality. Proektirovanie budushhego. Problemy cifrovoj real'nosti. Is.2. KIAM, 2019. P.228-249. [In Russian]