The blockchain was announced as the next digital revolution with multiple and practical applications of auto insurance contracts to the traceability of products in real time and intelligent data sharing .
Despite these beautiful promises, after 10 years of existence, the only application really deployed on a large scale, however, remains bitcoin with more than 20 million users . And again, this success is to be put into perspective, since cryptocurrencies account for less than 1% of the use of payment instruments in France.
Long validation time
Three factors can explain this slow diffusion: first, it is difficult to scale. The validation time of a block on the bitcoin network is 10 minutes . On the Ethereum network, we will remember the episode CryptoKitties , a fun application to raise virtual chats, which slowed the network for several days.
One of the main explanations of this long delay of block validation is the negative externality of mining. Indeed, the difficulty of mining increases with the computing power on the network. Thus, when a minor invests in new, more powerful hardware, he increases his probability of validating a block and thus being rewarded in cryptocurrency. However, it also exerts a negative externality on the other miners of the network who see the difficulty increase and thus their probability of validating a block decrease. They are then encouraged to invest themselves in new mining equipment, which increases the difficulty, and so on. Thus, the level of difficulty and therefore the time to validate a new block increases with the value of the currency, bitcoin or ether.
Secondly, the blockchain certainly brings elements of confidence, but at the same time poses new challenges. This trust is traditionally guaranteed by a sworn expert, such as a notary, who certifies that a transaction has taken place in accordance with the law. We talk about centralized trust. But even an expert can be wrong. For example, the Sotheby’s auction group was accused of having appraised a Caravaggio original as a copy .
An alternative is to turn to crowd expertise, such as the Wikipedia contributory encyclopedia, and online referral systems such as eBay. All studies show that the scientific quality of the Wikipedia articles is equivalent to that of the articles written by the experts of the Britannica encyclopedia.
Similarly, econometric studies show that eBay’s rating system quickly excludes bad sellers and allows good salespeople to increase sales and even sell more expensive products through a reputation bonus. However, this decentralized trust has the problem of public goods and the stowaway. Thus, on eBay, only about 50% of transactions are evaluated; on Wikipedia less than 1% of users would contribute between 77% and 90% of the content, according to estimates .
This is where the blockchain that guarantees distributed trust comes in: all the nodes in the network have a copy of the blockchain and thus contribute to its security. This is the undeniable asset of the blockchain: gather the two words “security” and “distributed”. Indeed, computer security is also a public good, both non-rival and non-excludable. As a result, the private sector tends to under-invest in IT security, as accounts of hackers regularly show. The blockchain solves the public good problem by giving direct monetary incentives when a network node contributes to the security of the blockchain.
But, as soon as you leave the blockchain, you open the possibility of hacking. This is particularly the case with the many intermediaries that have emerged to facilitate the management of cryptocurrency portfolios. Similarly, there are trading platforms that convert cryptocurrencies into other currencies. Private keys and other personal data are stored outside blockchain and are not subject to the same security protocol as in the blockchain itself.
There are hundreds of millions of losses in euros due to hacks (hacks) or because a user has lost his private key or has been extorted, or was stolen his hard drive. In total, one fifth of bitcoins are missing . Then, smart contracts , codes that trigger automated actions on a blockchain, especially on the Ethereum blockchain, use external data.
For example, a delayed flight insurance contract needs to consult an external source (called in the jargon a “software oracle”) before triggering an automated refund. If the input data is not reliable or if there is a bug in the code, the smart contract can have unexpected effects or even cause damage. This brings us to the third point.
Obscure aspects of governance
The governance of the bitcoin network communication protocol is crucial to ensure trust in the technology. This governance is problematic on two levels. First, the communication protocol is like a grammar for a spoken language. If we change the rules too often, we run the risk that people will no longer understand each other. The same goes for blockchain networks: we can not modify the rules of the protocol without consequences. Thus, when a major change is desirable, it is a question of ensuring that all the nodes of the network adopt the change of rule, because otherwise one risks a division of the network in two (“fork”), the nodes which accepted the change and those who rejected it. Now, and this is the second point, the computing power is now concentrated on a very limited number of farms and mining pools that influence the vote of a fork. These two relatively obscure aspects of the governance of a public blockchain can reduce the trust of users: forks risk dividing the network and consequently the value of associated cryptocurrencies.
To summarize, the blockchain solves the problems of trust in the digital economy but creates new ones through the problems of securing exchange platforms, smart contracts that require blockchain data and related governance issues at the forks.