Jean-Paul Delahaye is professor emeritus at the University of Lille, specialist in theoretical computer science and the complexity of algorithms. He just wrote Beyond bitcoin. In the world of blockchain and cryptocurrencies (Dunod, 270 pages, 19.90 euros), an educational book, but also committed, on the strengths and weaknesses of cryptocurrencies, as well as on their future.
Why is a mathematician like you interested in cryptocurrencies?
In 2008, Satoshi Nakamoto, whose name is still unknown, proposed a protocol, Bitcoin, which creates a currency independent of any central authority, whose operation is based on collective control, all being robust and secure. . The key idea is that of the blockchain, a kind of account book recording transactions without the possibility of erasing pages, which is shared throughout the network. It’s a brilliant invention. Before him, others had tried to develop electronic currencies, but without success. What is new and revolutionary is the development of a protocol that combines functions that are already known, but in an unexpected way, and that no one had imagined.
This protocol has also benefited from several advances to launch in 2009. Encryption, essential to operation, was quite mature, as shown by the success of banking transactions on the Internet. Similarly, peer-to-peer or distributed networks had been operating for several years without problems. Finally, as the system is based on the sharing of a large file of around 500 GB, the computers had to have enough memory and computing capacity.
However, you foresee the failure of Bitcoin, which you describe as a “diplodocus”, “Minitel of cryptocurrencies”, to be “stored in a computer museum”. Why ?
The main problem of the protocol lies in the way in which the validators of the new transaction pages to be recorded in the blockchain are designated. With Bitcoin, the chosen validator is the one who wins a calculation contest equivalent to proposing a kind of increasingly valid Sudoku grid, or rolling out a sixfold six with six dice… In practice, the calculation is linked to a cryptographic problem hard. The consequence is that there is a competition, increasingly tough and more and more expensive in energy. Initially, energy expenditure was low, then it increased tenfold approximately every year to represent, according to estimates, the equivalent of the annual consumption of Switzerland or Sweden, of the order of 100 terawatt hours! But it is a useless expense, because we can do without this type of method. That’s why I’m talking about a bug, an error for Bitcoin, which could have been better designed from the start.
You have 70.18% of this article left to read. The following is for subscribers only.