A neural network contribute to reverse cryptographic processes in bitcoin systems: attention on SHA256

Abstract

Bitcoin is a digital currency created in January 2009 following the
\nhousing market crash that promises lower transaction fees than tradi-
\ntional online payment mechanisms. Though each bitcoin transaction
\nis recorded in a public log, the names of buyers and sellers are never
\nrevealed. While that keeps bitcoin users’ transactions private, it also
\nlets them buy or sell anything without easily tracing it back to them.
\nBitcoin is based on cryptographic evidence, which therefore does not
\nsuffer from the weakness present in a model based on trust in guarantee
\nauthorities. The use of cryptography is of crucial importance in the Bitcoin system. In addition to maintaining data secrecy, in the case
\nof Bitcoin, cryptography is used to make it impossible for anyone to
\nspend money from another user’s wallet. In our paper, we develop the
\nidea that it is possible to reverse the cryptography process based on
\nhash functions (one-way) through Machine Translation with neural net-
\nworks. Assuming this hypothesis is true and considering some quantistic
\nalgorithms to decrypt certain types of hash functions, we will highlight
\ntheir effects on the Bitcoin system.