Fraud in IT - the trend has come to Russia or about the benefits of “Chimneys” (Due Diligence)

“It's hard to look for a cat in a black room, especially if he is not there”



While analyzing blogs and releases of familiar IT companies to understand the vector of development of the modern (including Russian) market, IT stumbled (yes, it was stumbled - this is not a typo) about an article in the Synesis blog (Synesis, Mink, Belarus). Link to the publication (in English).



Having contacted a company representative (one of the managing partners - Nikolay Ptitsyn), whom I personally know from meetings at various international events and parties (MWC, iGB Affiliate, etc), I found out a few details of the case.



Long story short: a certain young man by the name of Sancharov Kirill contacted the founders of the Synesis company (Minsk, Belarus) and proposed a partnership in the development of a mathematical data compression algorithm (archiver) using a completely scientifically-similar method of “searching for a given n-bit sequence in infinite an irrational number series, which is the value of Pi after the decimal point. ” The science-like sound, a combination of circumstances, as well as the stubbornness of the young “developer-entrepreneur” made fantasy a reality. A respected, experienced, multidisciplinary IT company picked up the project and gave the green light by allocating material, human and other resources for the development of the initiative and even provided its name for the presentation of the Aleph project at various specialized conferences and events (EMERGE 2019). 6 months of hard work of the company’s team and a desire to make a bold start-up business case broke up on obstacles of very strange properties, whether it were the periodic disappearance of the “author” of the idea under various pretexts (for example, “you must pass the session at the Moscow Institute of Physics and Technology”; at 26? Session? Didn’t I finish the 8th year otherwise?), Attempts to “hack” system servers by evil hackers, or the curvature of “developers” working on a remote site that periodically “broke the system” and needed an urgent personal presence in another country to fix the “damage” niy. " The demonstrated source code was suspiciously similar to the various generations of obfuscated PAQ solutions. Further - more, the data was magically compressed 2/4/8 /.../ n ^ 32 times. The apotheosis of history was the moment when the “algorithm” began to compress the newly compressed data the same n-times (though it is worth noting that for this it was necessary to encrypt the compressed data, otherwise the program recognized them and refused to “compress” again) and miraculously 1Gb imagine dense MPEG-4 data in a record 7Kb!



Synesis closed the project. The company estimates the total damage from the actions of a fraudster at hundreds of thousands of US dollars.



Sounds weird, huh? 2019, Python is taught in schools, IT has long become mainstream, and the number of people who understand the terms “data normalization”, “median”, “code encapsulation” amounts to tens, if not hundreds of thousands. Machine learning has not yet taught children (or is it already teaching?). In this environment, a situation similar to the one described above occurs.



Why? How? Is it true that the dissemination of knowledge leads not only to the growth of professional expertise in the industry, but also to the penetration of scammers whose lot was the theft of bank card data and the deception of vendors on electronic trading platforms?



I am inclined to admit that today 9 out of 10 professionals in the IT market (people who regard IT as their profession) have not even heard of Knut’s three-volume “The Art of Programming”. But logic is a stubborn science, and the theory of information has been studied by many involved in the industry! Following the logic and knowledge in the field of algebra, we can assume that with a certain amount of luck (the mathematical meaning of the term is meant by the author), as well as with the correct search / calculation / inversion technique, you can find a certain balance in the encoding speed / compression ratio and get some sane result. But the key word is balance. I immediately recall the triad of contradictions: “Quickly, efficiently, cheaply” - if you take any two assumptions as a premise, then the syllogism contradicts the third (“quickly and efficiently - it can’t be cheap”, “fast and cheaply - it can’t be qualitative”, “ high-quality and cheap - cannot be fast ”). Relying solely on methods of mathematical coding and not using the applied estimation of encoded materials, entropy cannot be defeated. The same “JPEG requires the application of a theory from digital signal processing, mathematical analysis, linear algebra, information theory, in particular, Fourier transform, lossless coding, etc.” (habr.com - citation, Philip Volodin @ Fil).



Afterword.



Any IT specialist is not new to hearing the term “due diligence,” but do we well understand the importance of such a familiar process in the real world? The story described above and the reason for this article leads me to rethink the importance of testing a hypothesis not only for its validity in terms of ideas and feasibility, but also in terms of checking the integrity of the “inventor”. “Grandma’s Algorithm”, which was a funny joke a few years ago, happened again - and this is an alarming sign!



The IT world is a world of chasing the “unicorn” in all its glory! And now, having caught up with something horny and seemingly of the right pink hue, the question arises: “Is it exactly him?”, “Is there exactly one horn (and the second is not cut down by sly people to idealize the result)?”, “And it’s exactly pink (in a deceptive binary light) the shades of the moon can change in the most bizarre way)? ”And other questions that we could come up with - they all add confidence to us on the one hand, and on the other hand, can be an unforgivable lag in making decisions.



I wish everyone simple and understandable dyudily, more ideas and real unicorns!