Many industrial, economic or social sectors, such as banking, logistics, agri-food or healthcare, see blockchain as an important innovation for securing and streamlining their digital transactions.
Initiated by Bitcoin in 2008, this technology is sometimes described as revolutionary, but remains often misunderstood and subject to justifiable doubts about its containment, especially its economic and environmental impacts. On September 15, 2022, the Ethereum blockchain, one of the main players in the field, made a major evolution, dubbed “The Merge”, by changing its internal validation mechanism, from “proof of work” to “proof of challenges”. “. Far from being a technical anecdote, this development removes an important obstacle that will make it possible to produce more efficient, reliable and sustainable solutions on an industrial scale.
The goal is to certify transactions, secure information systems and help establish true digital trust in a world where the digital modality of individual interactions has become essential and predominant. For example, the world of logistics and international transport in the major port areas is waiting for the roll-out of truly effective solutions that, in combination with the digitization of goods and container tracking, will make it possible to improve the port passage in a context of smart gate – a subject we are working on a lot in collaboration with the port of Le Havre.
But at the moment, the blockchain does not allow for high transaction rates (compared to payment networks with, for example, credit card), because it requires long calculations… which are also very energy intensive.
But what exactly is a “blockchain”?
A blockchain, or “blockchain”, is a decentralized digital ledger in which transactions are notarized. A network of validators act as digital notaries: each keeps a copy of the transaction log, preventing one of them from changing it without the others’ consent. New transactions are aggregated into chained blocks (see figure below).
Cyrille Bertelle and Claude Duvallet, Offered by the author
Each new block to be chained contains a “digital fingerprint” of the previous one, preventing any attempt to change a block: by changing the fingerprint, it would break the chain.
Cyrille Bertelle and Claude Duvallet, Offered by the author
The blockchain is qualified as tamper-resistant thanks to the two mechanisms described above: the chaining technique and the duplication on the network of validators.
Two questions then arise: first, how does registering a new block in the chain work so that it is accepted by all validators? How can these validators then be encouraged to work on the proper functioning of the blockchain?
The first question is based on a “consensus mechanism” among all validators – Ethereum’s recent mutation concerns this mechanism. The second question leads to the payment of the validator selected to register a new block with a cryptocurrency linked to the blockchain.
[Près de 80 000 lecteurs font confiance à la newsletter de The Conversation pour mieux comprendre les grands enjeux du monde. Abonnez-vous aujourd’hui]
The never-ending growth of resources needed for proof-of-work consensus
The first blockchain is Bitcoin, created in 2008 by Satoshi Nakamoto. The design is based on a consensus mechanism between the validators of the network, called “proof of work”, which allows to collectively validate each new block to be registered and to reward the selected validator for it. This one was selected because it is the first to solve a cryptographic problem that requires significant computing resources (computers, time, energy) that increase with the development of cryptocurrency.
At the beginning of Bitcoin, this “proof of work” mechanism was implemented thanks to the work of several dozen validators using microcomputers. Today it mobilizes several tens of thousands of validators working simultaneously (performing the same calculations to solve the same problem) on powerful computing resources called mining companies. The energy costs of these mining companies around the world are estimated in 2022 to be almost 75% of that of French households.
However, using Bitcoin is still relatively confidential compared to the rules of a banking network such as Visa or Mastercard. Its expansion would lead to ecological catastrophe before running out and then dying naturally unless it remains at an anecdotal level of use. In addition, related to the calculation costs, the transaction rates are very low (few minutes to make and validate each transaction) compared to those of banking networks such as Visa (thousands of transactions per second).
Ethereum, launched in 2015 on a new concept, that of “smart contract”, aims to secure and automate more sophisticated operations than simple exchanges of monetary values. This new concept thus aims to revolutionize the management of transactions for all the industrial, economic and social players mentioned above. In its early days, Ethereum used Bitcoin’s proven consensus mechanism (the “proof of work”), while announcing a move to other more energetically virtuous validation mechanisms.
Alternatives to “Proof of Work”
Consensus mechanisms have been the subject of much research before the blockchain appeared. After all, peer-to-peer networks that store and share data cannot function without these mechanisms, which validate the shared information.
Many conceptual models of consensus exist: proof of work (that of Bitcoin), proof of stake (at the origin of Ethereum’s “The Merge” evolution), proof of capacity, proof of delegated stake, proof of service, proof of authority , proof of trust just to name a few. This consensus is characterized in particular by the method of selection of validators and the validation mechanism itself.
The proof of stake, used for the evolution of Ethereum, is not based on the selection of a validator, after an expensive calculation performed simultaneously by all validators (as in Bitcoin). The selection is made by a random draw in proportion to the “stake”, which corresponds to an amount of tokens representative of the quality of its activity. Only the selected validator builds the new block — without solving costly cryptographic problems — and submits it to everyone else for approval. Thus, the cost of the operation becomes almost negligible compared to proof of work.
A limitation is that the choice of the consensus technique used should not lead to monopolies of certain validators (possibly supported by organizations or states) in an open and public system, thanks to their ability to compute or challenge. At the scale of a global network spanning many industries, Ethereum had to make sure it maintained system trust as it introduced its new consensus. While the “proof of stake” mechanism has been studied extensively, it has never been deployed on such a scale as Ethereum is aiming for.
In conclusion, public blockchains based on proof of work are doomed to fail for true sustainable implementation at industrial scale. Bitcoin is an example.
The migration to proof of stake (or some other more virtuous mechanism in energy costs) is essential for the blockchain to respond to the major economic, industrial and societal challenges it is expected to face.
But its deployment remains dependent on the need to reward validators on a public blockchain with a cryptocurrency that is inevitably linked to speculative appetites. It is necessary to know how to control this “edge effect” when engineering new information systems based on blockchains.