The greatest innovations result from new technological advances coinciding with a favourable sociological context that can transform these technologies into uses. In this sense, the blockchain is the product of, on the one hand, the convergence of asymmetric cryptography and distributed systems and, on the other hand, an opportune sociological climate. The latter is itself the result of a crisis of confidence from internet users towards their institutions, which has led them to seek out new forms of governance.
The rise of the internet has demonstrated how effective a worldwide communication system, free from the constraint of one unique, overarching telecommunications operator, can be. All around the world, it is now possible to connect to any given WiFi network in a matter of seconds.
Similarly, the blockchain is at the forefront of a revolution affecting the way we carry out transactions. It enables individuals to carry out operations among themselves, particularly those of a financial nature, which are guaranteed without the involvement of a trusted third party. This speeds up such interactions, and reduces their cost. Therefore, the existence of the blockchain is seriously challenging the role of institutions, banks and notorial studies, and having a profound effect on the way we approach administration.
The first experiments, which are by no means limited to bitcoin, such as decentralised autonomous organisations, demonstrate the radically disruptive nature of the blockchain.
Internet expert, co-founder of Sooyoos and Scanderia.
Member of National Academy of Technologies of France (NATF).
Definition of the blockchain
The blockchain is an innovative technology that enables users to execute transactions, of a financial nature or otherwise, that are guaranteed and can be audited by everyone without the need for a trusted third party.
After each transaction, a new line is added to the block, forming an indestructible chain: the blockchain. This is Accounting 2.0, with the history of each transaction indexed in a decentralised ledger and redistributed to all users. The complexity of the algorithms involved makes these transactions impossible to falsify.
‘La Caisse des dépôts lance officiellement l’initiative de place Blockchain’, caissedesdepots.fr, 31 March 2016
Government Office for Science, ‘Distributed Ledger Technology: beyond block chain’, 2016
Giulio Prisco, ‘Estonian Government Partners with Bitnation to Offer Blockchain Notarization Services to e-Residents’, 30 November 2015
Enguérand Renault and Benjamin Ferran, ‘Macron aménage la loi pour tester la “blockchain” sur la finance’, Le Figaro, 24 March 2016
Exaptation is the capacity of nature to create certain traits to resolve a specific problem, but which then serve another purpose entirely. See Serge Soudoplatoff, ‘Internet, entre percolation et exaptation’, in Martine Behar-Touchais, Nicolas Charbit, and Rafael Amaro (eds), À quoi sert la concurrence ?, Institut de droit de la concurrence, 2014, pp. 501-506
Please note: As often where digital matters are concerned, the subject of blockchains can get very technical. It is difficult to address this issue without providing certain technological insights, which can be found in the middle section of this document. Readers should not be afraid, for they will be perfectly able to negotiate these more challenging passages and reach the end of the document without losing their way.
Not a day goes by without the subject of “blockchains” cropping up in one form or another. They revolutionised money with bitcoin, and now they are poised to disrupt not just our banks, but our notaries, lawyers, estate agents, as well as the energy, healthcare, cultural and administration sectors. In brief, one would be hard-pressed to find an area of transactional human activity that will not be affected by blockchains. In the financial sector alone, banks have been experimenting with blockchains since July 2015, including institutions such as BNP Paribas, Société Générale, Citi, Deutsche Bank, Westpac, ANZ, Santander, ABE, DBS, Commonwealth Bank, UBS, Barclays, ING, Fidor and even the American Federal Reserve. The Caisse des Dépôts recently brought together sixteen separate institutions (four banks, four insurance companies, five manufacturers and three scientific partners) in a bid to better accommodate blockchains in France.1 The British government has published its own report;2 the Honduran government is trialling the use of the blockchain for its land registry, as part of the fight against corruption;3 the Estonian government is using it for notarial services for its e-residents;4 and the French Minister for the Economy, Industry and Digital Affairs intends to “develop regulations in order to test out the blockchain”.5
This is no longer about untested, futuristic projects. bitcoin was created in 1998, and its blockchain has not stopped growing ever since.6 Bitcoin cryptocurrency is based on an operational system that has fully proved its value, and whose construction required a lot less effort than other large financial transaction systems, some of which failed to become operational despite their enormous expense.
Although, to begin with, the blockchain was nothing more than the technology that underpinned bitcoin, it quickly became evident that it could be used for a whole lot more than just cryptocurrencies. To sum up, everything that is transactional in nature, be it financial or otherwise, can be put on a blockchain with the same guiding principle: to guarantee trust and better efficiency by, on the one hand, offering improved fluidity and a higher transaction rate and, on the other, by significantly reducing costs, via the simple elimination of the operational bottleneck known as the “trusted third party”. These changes are significant enough to be referred to as “disruptive”. Graphic 1 succinctly illustrates the diversity of what blockchains are currently used for.
We may legitimately wonder where the blockchain is currently positioned on Gartner’s famous “hype cycle”,7 and speculate that we are currently reaching the Peak of Inflated Expectations and that we will soon be diving headfirst towards the Trough of Disillusionment. Without falling into the trap of those consultancy firms that predict the future rather than shape it, what we can say for certain is that the blockchain story is only just beginning, that the twin mechanisms of percolation and exaptation that characterise the expansion of the digital world8 are going to set to work, and that the blockchain’s uses will ultimately prove to be much more varied and very different from those that we may imagine today.
Gletstalkpayments.com/blockchain-use-cases-comprehensive- analysis-startups-invovedraphique 1 : Possible uses for the blockchain
However, we still need to pay a lot of attention to the blockchain today. The blockchain not only represents a genuine shift, in terms of architecture, from the world of financial transactions (whose paradigm has not significantly changed since the invention of money and double-entry accounting) and other transactional domains. Indeed, and above all, it finds itself in a climate that is conducive to its growth, namely the current crisis of trust in institutions. It is exactly this combination – a new, powerful technological possibility in a climate that is conducive to disruption – that lies behind all the greatest innovations.
See Cévipof and SciencesPo’s ‘Baromètre de la confiance politique, vague 6bis’, published in February 2015
Cited in Paul Hoffmann (ed.), The Tao of IETF: A Novice’s Guide to the Internet Engineering Task Force, 2012
We are at the dawn of a veritable Renaissance. On the one hand, enormous progress is being made in science and technology. We are discovering exoplanets, exploring our own planet in increasingly fine detail and building quantum computers, just as during the Renaissance we invented the parachute, the dry dock and perspective in painting. We now know our place in the universe and are able to map it out with ever-greater precision, just as during the Renaissance we explored a world that suddenly had no limits. The internet is to our era what the printing press was to the Renaissance. We have the tools to understand our brains in more depth, enabling groundbreaking progress in neuroscience. We are sequencing the genome to such an accuracy that it may be divisible in units of 30 million, empowering improved diagnosis of illnesses, just as, during the Renaissance, André Vésale revolutionised medicine by challenging ancient Roman texts and dissecting the body with a modern methodology, reducing the personnel required from three people to one. However, just like during the Renaissance, the establishment is battening down the hatches to protect its privileges, refusing to change and killing innovation by demonising it, all in the name of holding onto power at any price. We are also currently going through a phase of economic regression which, fuelled by fear, is leading to a period of repression.
All of this has left a murky cloud hanging over the fundamental issue of trust. Transformation cannot take place without trust. Fear is a tool used by those in power who reject all change, and it is incompatible with trust. Who do we trust in 2016? Not Google, nor Facebook, whom we are entrusting with fewer and fewer of our secrets, nor Apple. We no longer trust brands, nor do we trust states. Even in France, one of the countries where levels of confidence in the State remain relatively high, this trust is ebbing away.9 Genuine trust exists today in only two domains: the family and the community. If war broke out tomorrow across the country, it is uncertain whether young French soldiers would stand up in defence of their nation, but they would undoubtedly defend their family and friends.
Trust is an unstable equilibrium. When two people trust each other, it only takes one of them to have doubts for the other to also start doubting. The result is that the parties descend into a state of mutual mistrust, a sentiment that is much less precariously balanced. It therefore takes energy to retain trust; yet it takes information to facilitate this energy. One of our era’s most violent breaks with established models concerns the source of this energy. France follows a model whereby energy is externalised: it is the nation’s judges, teachers, managers, parents, and so on, who are responsible for driving this energy. In the Anglo-Saxon model, the energy comes from both parties (or from the community, when there are several people involved). When eBay was created, it was not the only online auction and shopping website, but it invented the concept of buyers and sellers rating each other, a scoring feature that can now be found on all community sites such as Airbnb, BlaBlaCar, and so. What eBay understood is that trust could only be created by the community itself and not by the presence of third parties, which in its case would have meant expert auctioneers.
We could discuss the relative merits of the community-based trust model and the externalised trust model until the cows come home, but what we can say for sure is that the externalised trust model is proving inadequate in a world where the sheer volume of interactions is multiplying at such a rate, and it is starting to struggle. It is therefore very tempting for the regulator, the trusted third party, to demand ever greater resources in order to deal with this increase in the number of transactions. Unfortunately, this method clashes with the law of diminishing returns: past a certain threshold, the greater the means are, the more dysfunctional the system becomes. The community-based trust model is much more scalable and able to handle this increase in the number of interactions. Indeed, this is its main strength.
The blockchain model is even more powerful than the community-based trust model: it offers a model in which trust in transactions is reliable, can be audited by all and distributed thanks to its decentralised means of reaching a consensus.
Generally speaking, the construction of the internet was the product of a break from conventions. Whereas telephone operators were developing and maintaining a centralised network, the internet has shown the feasibility and, above all, the scalability of a totally decentralised network, without a centralised organisation and, therefore, without a single owner. Those same fundamental principles are there to be seen in its very construction: there has never been an “Internet project manager”, for the simple reason that the Internet has never existed as a project. The internet was built by “a loosely self- organized group of people who contribute to the engineering and evolution of Internet technologies”.10 Whereas the old world thought only in terms of broadcasting, and above all mass broadcasting, the internet has shown that everyone can create and deliver content – and that it is an error to try to apply the television model to the internet. Whereas the old world was based on “supplier to customer” chains, the internet has shown the feasibility of large- scale peer-to-peer exchange models.
It was inevitable that, at some stage, these new principles would be applied to the transactional model: whereas the old world believed in the necessary presence of a trusted third party, and whereas Internet 2.0 still features organisms offering platforms for interaction, the blockchain model shows that we can do without either and create a pure “peer-to-peer” model (P2P). In this sense, the blockchain is the transactional version of peer-to-peer networks such as Bit Torrent, which reflected – it bears repeating – the fundamental principles of the internet as far back as 1968, well before the invention of the Web (1991). The approach of this purely P2P model differentiates it from the “content provider” model (Web 1.0) and the “interaction platform” model (Web 2.0).
As in “data mining”, the idea of digging deep to resolve a For more information on this concept, see en.bitcoin.it/wiki/Mining.
This idea is down to Adam Back, the inventor of the Hashcash protocol, which is at the heart of the It is also used to guard against spamming. See en.wikipedia.org/wiki/Hashcash.
See Peter Sayer, ‘Bitcoin miner KnC is planning another four-week datacenter build-out’, com, 11 December 2015
The blockchain consists of two separate elements: firstly, a technology and, secondly, a system that uses this technology.
Historically speaking, the blockchain is the technology that underpins bitcoin. The invention of bitcoin in 2008 aimed to prove the feasibility of a currency based on a system of shared trust. It is an encrypted currency, whose trust mechanism is based on a system where the ledger is shared between multiple nodes of the network. The encryption algorithms for the transactions are open source, which reinforces this element of trust in the currency.
Bitcoin was effectively the first time that the trusted third party, a bank in this case, was demonstrated to be obsolete. Traditionally, it is the bank that guarantees the reliability and security of our transactions, making it the stereotypical example of an externalised trusted third party. Bitcoins are exchanged without being overseen by a trusted third party, but they still guarantee the same level of security, auditability and reliability. However, the subject of this paper is not bitcoin,11 but rather the technology on which bitcoin is based: the blockchain. The latter is itself based around three pillars: two are technological, asymmetric cryptography and distributed systems, and the third is sociological, the vision of a transaction model with a peer-to-peer structure, thereby enabling a distributed consensus to be reached without the need for a trusted third party.
The first pillar, cryptography, is based on the concept of a key. When symmetric, the key is held only by the two parties, and must therefore be secret; this has been known since ancient times. Asymmetric cryptography, which dates back to the 1970s, consists of combining a public key and a private key. The importance of this invention lies in the fact that it solves the problem of how to transmit a key without an intermediary. To illustrate how the mechanism works, we can use the example of a real estate property changing hands. Current conventions dictate that a notary holds the keys to the property, and oversees its transferral from one owner to another: this is a symmetric key. In an asymmetric transaction, the old owner places the key (along with any other possessions that are being transferred) in a trunk that can be secured by two padlocks. He sends it to the new owner having attached one padlock, to which he holds the key. The new owner receives the trunk and adds his own padlock, to which he holds the key. He sends the trunk back to the first owner, who removes his own lock and sends the trunk back once more to the new owner, who takes off the remaining lock, his own, and takes possession of the contents. On completion, all parties still have possession of their own locks, and nobody else has been able to open the trunk while it was being sent back and forth because there was always at least one padlock secured to it. There was therefore no need for a notary.
The second pillar is distribution. There is no finer demonstration of the feasibility of a distributed system than the Internet itself. Anyone, no matter where they may be in the world, can connect to the Internet, with no need for a unique, overarching telecommunications operator. It also bears repeating that the intelligence of the Internet is found at its extremities, and that the web itself is neutral. Nevertheless, the Internet is not “transactional” in nature; it is concerned with communication, and transactions are just one of its uses.
The third and final pillar of the blockchain – a distributed consensus – is an algorithm that offers us a solution to an amusing conundrum known as the Byzantine Generals’ Problem. Let us consider an army, split into several battalions, that is surrounding an enemy city. All of the generals must attack at the same time in order to take the city because, were they to attack separately, the city would be too strong. They must therefore find a way of communicating certain crucial information, namely the date and time of the attack, among themselves. Unable to meet in person, they must therefore send messengers. However, among the generals, there are traitors who may send false messages. For example, one of them may say to half his fellow generals that they need to attack at a given time, and to the other half that they need to retreat. This would lead to a difference in strategies, and the siege would be bound to fail. Until the invention of the blockchain, it was generally accepted that a consensus could only be achieved with the help of a central authority to coordinate all the generals: a higher power or, in other words, a trusted third party. The groundbreaking quality of the blockchain’s algorithm is to provide a way of reaching consensus without the need for this central authority.
The solution, discovered by the inventor of bitcoin, is as follows: each general can only send one order at a time, stamped with the time and date. But, most importantly, on receipt the orders are bound together and then encrypted, forming a chain of messages stored in a “register of transactions”, which is redistributed to all generals. In this way, if a traitor general receives the information “we attack tomorrow at 8 a.m.” and decides to only pass the message on to only half of the other generals and to send an order to retreat to the other half, he will create two incoherent chains. The honest generals at the source of the information will therefore be able to detect that foul play is afoot.
As such, a blockchain is an encrypted ledger, distributed and replicated at all the nodes of the web, containing order chains that are capable of generating trust without an external institution, thanks to consensus.
When a new transaction takes place, the information – including the time and date of the transaction – is placed in blocks of data that are integrated into the chain, hence the name “blockchain”. In order to be integrated, this chain is encrypted and must be validated. This validation work is carried out by the nodes in the network, computers capable of resolving the cryptographic problems required to validate any given transaction. This process of validation is known as “proof of work”. The people (or institutions) who perform this work are called “miners”.12
“Proof of work” is a cryptographic object that miners must present to prove they have spent a certain amount of time on the problem, which avoids simple cloning that would see a corrupt miner transformed into an army of clones.13 As the blockchain relies on consensus, it is important to guarantee those involved in it are “genuine participants”, thereby preventing any fraudulent retroactive manipulation of the blockchain. In fact, the mechanism is even more sophisticated than that: at regular time intervals, the level of difficulty increases.14
In order to encourage miners to validate transactions, they are set in competition against one another, with the first miner who manages to resolve the problem of a new block being rewarded (in bitcoins, at the start of this currency). In the early days of bitcoin, it was private individuals who carried out this work, through an innovative use of computer graphics cards, whose processing capacity is significantly greater than computer processors themselves. Computers then started appearing that were specifically designed to carry out this blockchain mining work. But with the size of the chains increasing, the processing power required has become enormous, and nowadays there are institutions that carry out this work. In March 2016, there were 7,420 nodes processing bitcoin blockchain around the world.15 New companies have started appearing on the market, offering mining services from their datacentres (“cloud mining”). However, the work is becoming less and less profitable for private miners.
The following photo shows a mining centre located in Boden, Sweden.16 It is important to note that, while it requires a lot of power to validate the blocks, checking an integrated block is simple and can be done by anyone. A distributed consensus is therefore easy to reach.
What the blockchain provides
The blockchain enables the construction of a vast ledger that is distributed as far and as wide as desired, visible to everyone, updated in accordance with a transactional principle similarly distributed and guaranteed by a community, without the need for a trusted third party as a central authority.
The blockchain makes five promises:
- Distributed trust.
- A system of transactions.
- Guaranteed by an extended community.
- No trusted third parties
- The capacity to operate complex protocols.
The blockchain is a genuine innovation: twenty years ago, it was by no means obvious that one day it would be possible for one technology to honour even the first four promises. Having said that, it is very much the combination of the five promises that defines the blockchain’s scope of application. If we needed a solution capable of fulfilling only one or two of these promises, other cheaper and more efficient methods would exist (see below).
The fifth promise is crucial, as it lends the blockchain its capacity for disruption: the ability to handle complex protocols (money transfers, banking, validation, and so on) in an automated way, with much lower transaction costs compared with systems that require human input, above all in the form of a trusted third party. In other words, the blockchain not only transports information, but also algorithms, and it does so with the same guarantee of trust as applies to the information itself. Already, the reader can begin to imagine the consequences that this could have regarding the automation of a whole range of processes currently carried out by human beings – notarised certificates, to take just one example. We will expand on this point below.
The key ingredients
Lest we forget, this has proved to be the main problem with the MD5 password encryption protocol: dictionaries are available online that allow users to search for the original word and, in turn, to break the key.
But there are still too many websites that store passwords in plain language, thereby creating dangerous vulnerabilities (for example lesechos.fr).
For the full calculation, see en.bitcoin.it/wiki/Scalability_FAQ#What_is_this_Transactions_Per_ Second_.28TPS.29_limit.3F.
The magic recipe of the blockchain contains five ingredients:
- Validated chains, that become almost impossible to falsify.
- Public and private keys that identify, and must be signed by, the participants.
- A peer-to-peer document distribution protocol (like BitTorrent).
- A large community, resistant to manipulation.
- A consensus validation protocol: the proof of work.
The understanding of this section requires an explanation of the role played by the “hash”. A hash is an algorithm that transforms a chain of characters (which may be a file) into a key, generally of a fixed length and which is hopefully unique, or at least possessing a low “collision rate” (a collision is when two chains have the same key). The hash is irreversible: it is not possible to recover the original text from the key, without the aid of a special dictionary.17 The hash is used to encrypt passwords: tests are carried out uniquely on the key and it is only the key that is stored on the database, which avoids the need to store passwords in clear.18
In the blockchain, the sequence of a fresh block and the key from the whole preceding chain is encrypted, providing a new key19 that enables the integrity of the entire chain to be verified. Once the key is established, no one block can be substituted with another, because the key would no longer be the same.
Let us consider the example of a corrupt individual who wishes to falsify data by changing, deleting or modifying an existing transaction. To do so, this person would need to recreate a whole new chain of blocks, starting from the date of the corrupted transaction all the way up to the moment the falsification takes place. More than half the nodes in the network would need to be convinced that the new version of the chain is correct. Thanks to the difficulty of proof of work, to do so requires too much effort in a limited time, as well as requiring the corrupt individual to possess over half the nodes. This is what makes the chain so robust. As with a public key, the cryptography work is asymmetric: it is very difficult to retrieve the original message using the key, but easy to verify the validity of the key, which in turn is what makes the blockchain easily auditable. Finally, another important factor is that the blockchain is not anonymous but pseudonymous: the participants in transactions are identified, even if their identity remains unknown.
Point 4 is also crucial. In order to do away with the need for a trusted third party, the community must not be open to manipulation. Point 5 guarantees the presence of a genuine community, but Point 4 dictates that the latter must be sizable and independent, so that no corrupt individual or institution can take control over the community of miners. This is particularly important because it means that the blockchain only takes on its full meaning on a large scale. In the same way as if there were only three Byzantine generals, there would be no guarantee of consensus, a small blockchain is useless. On the other hand, if it were just a group of trusting friends, there would be no need for the five ingredients: the first three would be sufficient, and the group could find a cheaper and more efficient solution than the blockchain.
Point 5 is the most complex since, as the chain keeps growing, the quantity of calculations required also increases. It works because the algorithm dynamically alters the cryptographic workload, and also because the size of the blocks remains constant. There is an ongoing debate over whether the size of the bitcoin’s blocks should be increased. As with all fundamental Internet decisions, this will be decided by consensus, with opinions currently being shared on a wiki.20 The mining time allowed to guarantee the proof of work, set at 600 seconds per block, also remains constant. In practice, there appears to be a limit of 6.6 transactions per second.21 The good news is that this works on a worldwide scale; the bad news is that, in order for it to work, a significant share of the world’s processing power is being used up.
Wonderfully enough, in its early days the blockchain used processing power available on empty machines that were distributed throughout the community, which was doubly beneficial in terms of both CO2 emissions and Point 4. At the same time, private individuals used graphics cards, which were more powerful than their computers’ actual processors. There are also specific “blockchain mining” machines available for sale over the internet. But at the moment, 50% of all mining power is in the hands of a limited number of Chinese entities, who are using specialized hardware.
What we can do without the need for the five promises
APIs are programming interfaces that allow an external information system to be They can be considered as supply points that deliver specific services.
Fulfilling promises 4 and 5 comes at a price. Indeed, so long as the credibility of the trusted third party is assured, the same outcome will generally be possible at a fraction of the cost. For example, a trusted third party can offer to register and validate documents and information in a digital depot, set up in the form of two application programming interfaces (API).22 The first API allows anyone to store information, and the second to validate a claim. Promises 1 and 2 are fulfilled by this transactional approach, and the capacity to distribute trust is carried out in the form of the verifiable certificates. Although we must accept that the depot itself is a black box, it is easy to make and a lot less expensive than a blockchain.
Promise 3 stipulates that the depot should be open and visible to a large community. Promises 1 to 3 can therefore be fulfilled with a blockchain that is shared with all but validated by just one (the trusted third party). In this eventuality, the structure of the distributed blockchain is of real interest: a chain is a way of incrementally guaranteeing integrity; furthermore, by using a peer-to-peer distribution mechanism, many participants are empowered to check the coherency of said chain. In terms of API, this means that the first step of the transaction is retained (insertion, followed by confirmation of receipt by the trusted third party), but that the verification API is no longer required as the chain is widely distributed and can be checked by anyone, as is the case with the blockchain. This solution satisfies promises 1 to 3, and such an approach is therefore armed with one of the disruptive powers of the blockchain (“trust as a service”), from the moment trust is placed in the third party.
Solutions do therefore exist that envisage the production of (much cheaper) alternative blockchains that can be applied to functional sub-domains. As such, the main interest is to reduce transaction costs by eliminating human work that can be carried out in greater security and at a lower cost by algorithms. For example, by using a blockchain, transferring money from one country to another would cost a tenth of the price and would take ten minutes to be validated, instead of several days. It is this promise that has suddenly caught the attention of the financial sector.
Let us remind ourselves one last time: the blockchain’s claim to be resistant to fraud does not hold up when confronted by a declining community. In other words, a small community cannot protect itself against a hostile third party possessing very strong processing power. This is why the blockchain requires global interest that can attract a community of miners from all over the world. The questions that need to be checked carefully are whether it is possible, for every proposed use of the blockchain, to assemble a global community around the subject (without which Point 4 is not met and there is a risk of submersion) and whether the cost of the huge proof of work remains cheaper and more fluid than the services of a trusted third party.
From a political point of view, the response may be very different (here it is not cost which is at stake but liberty, and this is what motivates the majority of the crypto-blockchain community). But from an economic point of view, potential uses are still emerging. In the world of traditional finance, the idea of doing away with the trusted third party is unthinkable. Any initiatives currently emerging from among the financial institutions do not respect all of the blockchain’s promises; if they did, the cautious and reserved signs of interest that banks are showing in the blockchain would surely be replaced by a wave of panic, since the blockchain fulfilling all of its potential would render their role of trusted third party obsolete. On the other hand, when transaction costs become too high, in terms of both time and money, the credibility of the trusted third party evaporates. Under these circumstances, the third party-free blockchain offers a more efficient alternative.
The situation somewhat differs when viewed from a social or political perspective. When citizens genuinely start challenging the efficiency of administration services, the latter should look into the blockchain. There are numerous examples of situations in which the blockchain can be of real political or societal benefit, and in which we could happily bypass our institutions, or even the State. As such, a new model is emerging known as Decentralised Autonomous Organisation (DAO), described in greater detail below.
Three key issues
As a reminder, 1 peta = 1015, or 1,000 tera
The DNS is the Internet’s global directory. It is the DNS that decides that a given address (www.fondapol.org, for example) will direct users to its corresponding server. If it ceases to grow outwards the Internet will explode, as the same address would direct users to other servers depending on their location. This would signal the end of the web’s worldwide nature.
The first important issue is that of latency. Due to the way the blockchain is constructed, two factors directly impact on its latency: firstly, the proof of work requires time and, secondly, the validation of each transaction block depends on the requisite probability of it being secure. Therefore, a certain amount of validation time is needed, which fluctuates randomly like a queue (the validation time is at least 10 minutes per block for bitcoin). Any scalability problems essentially translate as slower response times. There are many financial services that will not tolerate such delays, above all at high frequency trading times.
“Scalability” is a more complex issue, and the subject of much debate. In order for the blockchain to continue to grow harmoniously, an ever-increasing level of processing power is required (measured in petahash per second),23 but the “distribution” of the community must be preserved (i.e. a significant number of independent participants). Ingredient 4 effectively states that the community of miners has more resources at its disposal than a malicious attacker, and that the community is sufficiently vast and distributed so as to resist any attempts to take control of it. However, rapid growth and the need for scalability mean that the community is becoming increasingly concentrated. An interesting parallel can be made with the distributed architecture of Domain Name Systems,24 which is also experiencing a de facto concentration.
The phenomenon of distributed trust also generates costs. In the early days, the processing power consisted of free cycling power provided by under- utilised machines. The competitive nature of the consensus process (the first users to reach consensus take a share of the reward) created a Darwinian environment. The energy cost of the proof of work process is not negligible, despite the progression towards specialist ASICs, a trend that, furthermore, goes somewhat against the preservation of an open community of developers. We are starting to hear complaints that the validation costs are becoming high in comparison with other more traditional methods. This is only going to get worse: the rules of Moore’s law go in favour of those other methods (the costs of which will continue to fall as machines become more powerful at less expense), while the blockchain, by its very nature, will require more and more effort as computing technology advances. Alongside proof of work – in which, let us be reminded, everyone must solve the same puzzle – another means of reaching consensus has emerged: “proof of stake”,25 where the miners’ effort is concentrated only on the subsets of the blockchain that they own. Bitcoin is based on proof of work, but other cryptocurrencies have tended to move towards proof of stake, such as Peercoin.
Philippe Ricard and Patrick Roger, ‘TVA : 32 milliards d’euros perdus par la France chaque année’, lemonde. fr, 18 September 2013
The principal manifestation of this idea is nicely explained in a video entitled ‘Rent, sell or share anything – without middlemen’ (slock.it), produced by a start-up that enables its users to rent out or lend any personal belonging, thanks to a blockchain
See ‘Airbnb just acquired a team of bitcoin and blockchain experts’ qs.com, 12 April 2016
Anyone who has ever read a contract will be able to attest to how complex they are. Lawyers seem delighted by the prospect of creating such complex documents, which nevertheless makes them difficult to execute and increases the risk that they contain contradictions. But it often feels like there is a whole series of hoops to jump through to establish even the simplest of contracts.
In 1993, the concept of the “smart contract” was invented to automate contractual relations, by eliminating human intervention. A bank loan, for example, is perfectly capable of being entirely automated, without any human input, since all of its conditions are impartial. But there will always be doubts over whether a contract is being correctly executed; in other words, over the auditability of the contract.
What makes blockchain technology unique is that it enables the storage of not just content, but also algorithms, thanks to the sections of code that it holds. As the blockchain enables everyone to audit these algorithms, trust can only be strengthened.
Let us consider a simple example: VAT. There are enormous sums to recover in unpaid VAT (€32 billion in 2013, in France alone)26 and fraud, above all so-called “carousel fraud”, is responsible for a large share of these missing payments. Now imagine that all VAT transactions were stored on a blockchain. All the parties involved could carry out their own audits in order to ensure that the rules were respected and that all transactions were entirely above board – that calculations were correct and that payments had been made. Fraud is therefore no longer possible.
The verification costs are a lot lower than if the checks were carried out by a human operator and it is, above all, a lot quicker. This remains true, provided the events that enable those involved to verify the execution of the contract can be automatically detected by the blockchain, like in the case of a bank loan and its repayments. But what if, for example, an action (such as a payment) were to be triggered by a physical event, like the delivery of goods? This is where the “smart” elements that make up the smart contract take on their full meaning. Imagine a world of electronic keys, where transferral of a real estate asset would be automatically triggered by the execution of a sales or rental contract present in the blockchain. The contract would be impartial and fully auditable, which would in turn make the property inviolable. The former owner would no longer have access to the property because the old electronic key would no longer work, and the new owner would only be granted access once the software had unblocked the new key.27 Airbnb’s recent decision to experiment with blockchain technology could well mark the start of a new trend of tenant-owner relations being governed by smart contracts.28 Only now is it becoming clear how the professions of trusted third parties such as notaries, lawyers and clerks could be totally transformed by the blockchain.
Autonomous decentralised organisations
For more on the importance of a society based on knowledge, see Idriss J. Aberkane, Économie de la connaissance, Fondation pour l’innovation politique, 2015
Tim O’Reilly, ‘Government as a Platform’, in Daniel Lathrop and Laurel Ruma (eds), Open Government. Collaboration, Transparency, and Participation in Practice, O’Reilly Media, February 2010, chapter 2
See boardroom.to. Their white paper nicely sums up the workings of a DAO: Nick Dodson, BoardRoom: A NextGeneration Decentralized Governance Apparatus, n.d
The world of business, just like the world of administration, is experiencing a fundamental crisis. The reasons are the same in both domains: silo business models, vertical structures and overbearing hierarchies; management based on mistrust; governance leaving little space for creativity and invention; and a differentiation between the “thinkers” and the “doers”. In a world of interactions, where collective intelligence is the rule, these models are inefficient because they do not sufficiently circulate information and develop knowledge.29
In 1937, the economist Ronald Coase showed that the concept of the firm was chiefly designed to reduce transaction costs by, among other methods, connecting information and logistics.30 In this model, hierarchy is important because it reduces uncertainty and, in turn, transaction costs. But why is the world not, therefore, one big firm? Because a second cost must be added to the first: the cost of organisation. In other words, the law of diminishing returns dictates that the benefit is not always proportional to the amount invested. The peer-to-peer model enables the ongoing transmission of information while guaranteeing trust. This also applies to the transaction model.
Ronald Coase was also interested in social costs. He showed that the State does not have enough information to impose all taxes at the correct level, but that tax agents and taxpayers could come to an agreement – in a “peer-to- peer” mode, to use today’s parlance – so long as the transaction costs were low.
There was just one missing link to make Coase’s ideas workable, and it is blockchain technology that can serve to bridge that missing link. We can now see the extent to which the blockchain is seriously undermining the principal raison d’être of our institutions.
An organisation is made up of tangible assets, intangible assets and people. According to the traditional paradigm, certain people make the decisions (the board, management teams, the parliament, the government) and others execute them. The industrial revolution significantly reduced the number of “doers” and replaced them with robots. But white-collar workers will soon be going the same way: brainpower may have replaced manpower,31 but it too can be automated and replaced by computers. The more a firm is governed by rules and processes, the more obvious it will become that they should be automated. If we return to the example of the bank, the presence of a “middle man” is by no means required to execute a money transfer. The latter slows down the process and contributes nothing, only becoming useful when it comes to bending the bank’s rules. In order to make a valid transaction, a company’s rules can be applied via software alone, as we have seen with smart contracts. This does not signal the end for the human being, of course, but rather marks the emergence of a business model that will no longer use our intelligence to carry out repetitive tasks that bring no added value. Instead, our intelligence will be used to create knowledge. All the conditions are now in place for us to create an entirely efficient business model, where all stakeholders can participate in decision-making, can audit the rules and can check that they are applied. The political equivalent of this phenomenon would be “government as a platform”, as defined by Tim O’Reilly.32 And the blockchain is the tool that enables the management of these organisations.
DAO is a theoretical model of governance whereby autonomous entities cooperate with each other in accordance with an unfalsifiable set of working rules. To achieve this, one method is to implement the rules by using open source software distributed onto the computers of all stakeholders. A sample set of encoded rules of governance can be found on the Ethereum website.33 Seeing the rules in their code form may seem bizarre, but it makes them easier to understand and, therefore, easier to audit. One side effect of using a blockchain would be to verify the consistency of the rules of governance. The codes are sure to be full of contradictions, which become far easier to detect. Another interesting effect is the possibility, via the blockchain, of implementing liquid democracy, in which each person can choose a representative to vote in his/her place for certain decisions, within a limited time and space. For any readers interested in this model, a start-up named Boardroom offers DAO-specific management tools.34
AltaVista existed before Google, but was unable to find the right economic When Digital was valued prior to its acquisition by Compaq, AltaVista was not even taken into account.
Giulio Prisco, ‘Microsoft Launches Ethereum Blockchain as a Service (EBaaS) at Devcon, Boosts Ethereum’, 11 November 2015, bitcoinmagazine.com
In the world of the Internet, very often the first party to arrive takes all the spoils, providing it finds the right economic model, what is called “winner takes all”. We saw it with Google and Amazon, then Airbnb, Uber and others.35 With respect to the blockchain, only one company is currently emerging that offers a generic blockchain: Ethereum.
Ethereum consists of a foundation based in Toronto and a company based in Switzerland. It offers a blockchain that enables users to manage not just cryptocurrency but also smart contracts, via a Turing machine. Its code is open source and its currency, which is called ether, was worth $900 million in April 2016.36
Ethereum works on several levels – simultaneously identifying what must be done to mine and what is possible to validate – as the ultimate aim is to have a complete validated Turing machine. There is a risk of amassing such a high level of complexity that other problems emerge, at a time when we are still a long way from having explored everything that it is possible to do with the blockchain. But we must keep faith in the capacity of the Americans to resolve any problems that may arise.
Microsoft has just started offering “blockchain as a service”, based on the Ethereum technology.37 It remains to be seen whether Ethereum will become the next quasi-universal service, or indeed whether Amazon will pursue its Amazon Elastic Compute Cloud (EC2) project and create its own blockchain.
The blockchain is already used in numerous different ways, and within varied domains. It is no longer possible to list them all here, but it is nevertheless possible to give some examples. In all likelihood, an overview of the uses in a year’s time (in 2017) will look entirely different to how it does today. On the other hand, not all of these experiments truly fulfil all of the five promises. It is hard to imagine a trusted third party suddenly creating a blockchain that honours promises 1 to 5, without first considering how its new role will look in a world where the blockchain guarantees trust in all transactions, thus rendering the third party obsolete.
The first example of the blockchain in use was, of course, the bitcoin. This cryptocurrency, which respects all five of the blockchain’s promises, was
invented in 2007 by the mysterious Satoshi Nakamoto. It is limited in quantity (21 million) and is starting to be a prominent fixture in the landscape: in May 2016, over 7,600 destinations around the world accepted bitcoin as a method of payment,38 and the US Securities and Exchange Commission (SEC) has even authorised donations to political parties made in bitcoins. Bitcoin has quickly been caught up by other cryptocurrencies: Wikipedia numbers them at over 600, including 9 with a value of over $10 million.39
Finance is a typical example of a model that finds any sort of change difficult. Not only is the cost of banking transactions enormous, but they are not at all fluid: we still have to wait several days to carry out an intra-European money transfer, in return for a service that is not of great quality. And above all, banks are very reluctant to open up; it took the introduction of PayPal for them to start opening their APIs. Any human intervention in a transaction slows it down, resulting in a lower overall processing capacity, and therefore a poorer quality of service. The cost of mistrust is immense. One of the blockchain’s great assets is that it empowers the progression to a model based on trust.
For customers, the blockchain’s great strength is that it speeds up transactions while preserving collective trust. For financial institutions, the blockchain represents an enormous reduction in costs and the possibility to offer a better service. But a blockchain that fulfils promises 1 to 5 essentially renders the institution obsolete. This is why banks are currently in the process of building blockchains that do not honour all five promises, with a view to reducing costs and making transactions more fluid. Right now, these traditional financial institutions are only at the start of their experiments. But Estonia, for example, decided, with the help of Nasdaq, to put the voting procedures for the shareholders of every company in the country on a blockchain.40
Our healthcare system dates back to ancient times.41 There is barely any information transferred between the various stakeholders (town doctor, nurse, hospitals, and so on) and it is still the patient’s job to take care of communication by bringing his/her own medical file to appointments. France’s
personal medical file project (known as the dossier médical personnalisé) has been a dismal failure, for purely political reasons. The result is an increased level of suspicion in healthcare institutions.
A blockchain would offer many benefits: first and foremost, there would be no more trusted third parties wasting the public’s money on complex systems that do not work. The blockchain can help to create a healthcare system whose construction and running costs are lower, thus increasing the amount of money available for patient reimbursement. Furthermore, the ability to integrate smart contracts into the blockchain enables a much more personalised service, with expenditure and reimbursements calibrated in accordance with each individual profile. Ultimately, the community would be rewarded with a fully functioning service, rather than opaque administration.
This is not a utopia: Estonia, a country renowned for investing heavily in digital solutions, is currently creating a blockchain to store the medical files of all of its citizens.42
See Aaron Wright and Primavera De Filippi, Decentralized Blockchain Technology and the Rise of Lex Cryptographia, abstract, Social Science Research Network (SSRN), 10 March 2015
Dominik Schiener, ‘La démocratie liquide : une véritable démocratie pour le 21e siècle’, n.d
The complexity of French law and regulations is not a matter open to debate. Labour laws alone consist of somewhere between 2,000 and 15,000 pages of text, depending on what we consider to be core regulation and what to be jurisprudence. But even 2,000 pages is an awful lot, particularly when we also consider the branch agreements, special status regulations, European laws, and so on. And let us be quite clear about this: no politician will ever have the courage to reduce the French labour law. It would be a Herculean task. However, it would surely be beneficial to codify these texts into smart contracts – and in doing so iron out any contradictions held within them – and then to place these smart contracts in a blockchain, which would then be shared with all stakeholders: companies, administrations, employees, etc. All the calculations would be automatic and the financial gains for the state, and therefore for all involved, would be enormous in terms of control.
Generally speaking, all rules of governance, be it for a company, a charity or a country, can be put on a blockchain (as detailed during the section on DAOs).
In politics, this would result in the concept of “liquid democracy”. Debates are currently raging as to the limitations of this system, which could lead to the “tyranny of code”.43 Certain political parties, like the Pirate Party in the UK and Nous Citoyens in France, already use blockchains to manage their voting procedures.44 One political party in Australia, The Flux Party, decided to build its governance on a blockchain. The principle is that the party’s senators must apply the decisions chosen by the members’ vote, which takes place via the blockchain.45 The innovation is that members each have voting credits, which they can exchange on the blockchain in order to focus their votes on their own personal areas of interest. This is a genuine example of the principles of liquid democracy in action, with delegation on certain subjects and direct voting on others.46
OpenBazaar (still in beta) is a 100% peer-to-peer classified advertisements platform.48 Instead of having to visit a website, users download software onto their computer, which they use to access what is on offer or to sell their own items, without any commission. It is a competitor of eBay and the French website, Le Bon Coin.
Collaborative transport also has its own blockchain. Lazooz coordinates a journey share service and, of course, all financial transactions on a blockchain.49 Just like Open Bazaar, the main benefit is to reduce transaction costs. However, it still remains to be seen whether the presence of a trusted third party remains necessary for the success of a car sharing service, as it gives the user a partner to turn to and share the risks, thereby guaranteeing customer satisfaction.
For more detailed information on the sidechains mechanism, see Adam Back et al., ‘Enabling Blockchain Innovations with Pegged Sidechains’, abstract, 22 October 2014
There is a tension that pits the diversity of opportunities discussed above directly against the need for one unique, worldwide infrastructure that guarantees the community remains larger than any potential attackers (see Ingredient 4). We have seen that the idea of having “my own little blockchain, all to myself” is not compatible with the five promises. However, once a trusted third party is accepted, it is nevertheless easy to instantiate a whole subset of these technologies for individual cases. For example, for the certification of documents (land registry, damages, property, etc.) a simpler system, implementing promises 1 to 3, is sufficient, and does not entail the additional financial and energy cost related to proof of work.
Nevertheless, using the blockchain to create a trust as a service model makes a lot of sense. Essentially, the beauty of the blockchain approach is to enable a small unknown entity, for example a start-up, to offer the same guarantees of transparency, sustainability and other trust-related characteristics that are traditionally associated with established, institutional structures (this is the competitive advantage held by large financial institutions). Once it starts registering its transactions in the international blockchain, this start-up will offer a non-repudiation guarantee that is the equal of, or superior to, that offered by a State or a bank. It nevertheless remains unclear whether the current infrastructure is able to host the avalanche of requests and opportunities that we have briefly alluded to here.
Consequently, we are witnessing the emergence of a tree structure: a large central blockchain, available worldwide and validated by a vast community, with branches (blockchains or otherwise) operated on a simpler level by start-ups or small communities with an interest in them. The interest here is that the start-up can place the ledger of its own activities in the central blockchain, immediately making it trustworthy and transparent in the eyes of its customers. The ledger, meanwhile, can be managed with lighter techniques that require less of an investment in terms of time and money.
This approach has given rise to several technological developments, including sidechains. A sidechain is a chain of transactions managed by a sub-community, with similar encryption and authentication techniques as the blockchain, but with a simpler protocol facilitating improved performance levels. The distribution of control (the sidechain is controlled by a smaller group) lends it more agility, but the end of this sidechain (the peg) is integrated within the blockchain so that the former benefits from the increased security of the latter.50
This solution also extends the blockchain with the addition of richer protocols, meaning that the “blockchain/sidechain” blueprint is a better direction for the ecosystem to evolve in than the creation of new, autonomous blockchains.51
Can a true peer-to-peer mode survive without the presence of a large entity behind it? For Uber or Airbnb, the brand value is not in the platform but in the promise made to their customers. And it takes human resources to provide the service once the sale has been made. But on the other hand, the Internet itself functions without the presence of such an entity. Early internet detractors pointed towards the absence of an operator, which supposedly rendered it too inaccessible for the uninitiated. This is true, but community help systems have worked perfectly, and have in time replaced and improved upon the much less efficient hotlines and call centres.
In the future, when the blockchain’s influence becomes truly disconcerting for the established institutions, there will be a great temptation for those currently in charge to suppress it, by outlawing it or limiting its effects. The Internet was born in 1969, but did not become widely available to the public until 1991. It is only 25 years later that the majority of politicians are now trying to suppress the innovation that the Internet brings with it.52
But attempting to hold the Internet back is like trying to stop the rain. The decentralisation of the web, the fact that intelligence is found at its outer limits rather than inside the network, the longing of many citizens for another model where they are more engaged, and above all the growing complexity of the world, characterised by an increasing number of interactions, will encourage our progression towards distributed trust. Any human intervention in a transaction slows it down, resulting in a lower overall processing capacity, and therefore a poorer quality of service. The blockchain’s great strength is that it speeds up transactions while preserving collective trust, all at a lower cost.
Thanks to the invention of the Internet technologies, the world of telecommunications has progressed from a centralised model with a prominent role for trusted third parties (the operators) that justified their role by the promise of “total quality”, to a decentralised model where everyone can effortlessly connect to the internet from anywhere, and benefit from a universal array of low cost services. Thanks to the invention of the blockchain, it is increasingly likely that the world of transactions, and not just the world of finance, will experience the same disruption; one that will not prove any less painful for the operators.