Blockchain: not the Messiah

Blockchain is an ingenious technological development, but it's far from being the saviour many are looking for

The word "disrupt" is never far away from any mention of blockchain; "revolutionise" puts in a regular appearance too. And so does the word "save", both as in saving money ("Here's how blockchain will save global trade a trillion dollars") and also as in removing an existential threat (Why blockchain will save the music business in 2016). Disruptor, revolutionary and saviour - blockchain's certainly got a lot on its to-do list.

As with the hapless hero of Monty Python's Life of Brian, people seem determined to read into blockchain virtues that aren't really there. They extrapolate its ability to enforce the transparency of certain transactions to imagine a world in which all transactions are transparent. Hearing tell of its decentralised nature they concoct a promised land rid of corrupt and venal central authorities and chiseling middlemen. The tables are turning and the money changers are about to be expelled from the temple, they prophesy.

But blockchain is not the Messiah. It's not even "the" blockchain (who knows where that self-important definite article came from?). It's just plain old blockchain, a very interesting idea, the basis of new currencies that could (but haven't yet) alter money forever and an important technological development in its own right, but one that's flawed just like the rest of us.

Keeping us honest

One of the key virtues of blockchain is its ability to keep us honest as we transact with one another by creating a permanent, incorruptible and immutable record, the veracity of which no one can dispute. Much of the interest in the technology derives from this potential to do away with the trusted (but possibly untrustworthy) intermediary in a transaction, be that a bank or a lawyer or some other agency.

Blockchain is an example of a mutual distributed ledger (MDL). All transactions written to it are visible to all, shared across multiple locations and impossible to alter or delete. As the basis for establishing consensus and enabling contracts between multiple organisations with no intermediary this is extremely useful. It also creates a permenant audit trail - imagine how financial regulators might be helped if the entire audit trail leading to the 2008 financial crash were right there in one place.

Blockchains could potentially be used to eliminate fraud, not just in financial transactions but in the supply chain too.

"Using blockchain you can stamp each transaction, each shipment," says Gareth Stephens, head of proposition development at identity data intelligence company GBG UK, citing the example of the diamond trade.

"The individual biometrics of the diamond get stored in the blockchain so you could look at the provenance and make sure it's come from a trustworthy mine and that it's not a blood diamond."

Yes you could, in theory. In theory you could also buy a house in a day, but given the number of intermediaries who make a living from this process this is unlikely to happen. How hard would it be to get all the middlemen and merchants in a diamond supply chain to sign up to the same system, one that might end up costing them dear? "It's probably 10 years away," Stephens concedes.

There are other practical limitations too. Blockchain is a tool for recording transactions, not for authorising new transactions, and transparency of record takes no account of what is ultimately done with this information - that's down to the lawyers and politicians who will have to agree (on a global basis) exactly what constitutes a verifiable contract in a blockchain world. Even then there is no guarantee that it will improve justice. There is plenty of evidence who the guilty parties were in 2008, but that doesn't mean they got punished.

It's also not hard to imagine instances in which the permanence of information on a blockchain could be a problem once uses are rolled out that go beyond the relatively straightforward financial transactions and authentication cases - online voting for example. Blockchain transactions can already be "reverse engineered" to allow a pretty good guess at the transactor. Will the encryption algorithms remain sufficiently robust to protect these records forever, or will public blockchains become a gigantic honeypot, a juicy target for states and hackers to attack? Permanence of record and the end of plausible deniability can be a double-edged sword.

Blockchain: not the Messiah

Blockchain is an ingenious technological development, but it's far from being the saviour many are looking for

Designed by anarchists

Another blockchain virtue is decentralisation. Centrally stored repositories of sensitive data are inherently insecure - just ask the customers of Ashley Madison, Target and TalkTalk.

"There's a massive requirement to decentralise these customer databases and blockchain just plays perfectly into that," says Stephens. GBG UK is working with blockchain to develop a federated identity infrastructure that allows individuals to identify themselves online without having to supply a driving licence number or credit card multiple times to multiple organisations.

Rather than being held in any number of central databases, sensitive data is encrypted and distributed and a blockchain is used as an audit ledger to re-authenticate the user, to make sure a person who is logging into a website is who they say they are. Once in place, Stephens says this will mean the end of the password. The user will simply click on the logo of a chosen identity assurance provider (e.g. GBG) and the login process will be taken care of by matching encrypted hashes held on their device with those on the blockchain. The blockchain will also be the open platform on which rival identity assurance providers can compete, reducing the risk of monopoly, Stephens says.

Similarly, a blockchain could be used by the NHS to enable patients and practitioners to manage electronic patient records (EPR) in a more secure, consensual way. Gareth Baxendale, vice chair of the Health Executive at BCS, the Chartered Institute for IT, points out that one of the problems with the care.data scheme was that it involved centralising patient records in a way that made many people uncomfortable.

Using a blockchain could allow the NHS to retain the current decentralised model in which EPRs are stored locally. The blockchain would store metadata, rather than the EPR. That metadata could be used by a medical professional together with the patient's unique ID to update the EPR from anywhere in the world.

"This metadata would use a hash value in the blockchain, demonstrating data existence and confirming the data integrity without revealing the actual data itself, which is a key requirement in the world of patient records," Baxendale says.

But is decentralisation necessarily a good thing? Not always, says Tom Loosemore, formerly CTO at UK government digital services (GDS), who says that ultimately we have to place trust in institutions.

"We looked at blockchain [for managing registers, the repositories of trusted government data] but we didn't like it for a number of reasons," he says. "One is that it's fundamentally designed by anarchists who don't want there to be a central authority at all... but in a democracy you do trust and appoint people to be in authority."

Can you keep a secret?

Blockchains record and retain every transaction taking place on them, but they don't record who the exchange is between. That's because addresses (such as a bitcoin wallet) are hashed. But that doesn't make them anonymous; rather they are pseudonymous.

There are companies out there such as Chainalysis that can join the dots by analysing the traffic between addresses. If you frequently trade with one or two addresses it will not be too hard to guess your true identity. The business of these companies is fraud and crime prevention rather than snooping, but their existence demonstrates that activities on a public blockchain cannot be considered confidential. If parties wish to keep secret the fact that they are in communication - during a corporate buyout perhaps - they should probably steer clear of blockchain. A surge in traffic between two addresses could reveal a lot more than the participants ever intended.

A solution to this is a private blockchain, where only certain approved users can conduct and verify transactions. However this immediately nullifies many of the transparency and decentralisation benefits of the public blockchain model. The question of how restricted such a solution can get while still being called a blockchain is the subject of some semantic debate.

Small private blockchains are the way things are going with the world of smart transactions, though, in which a washing machine can order its own washing powder without human intervention. More than likely we'll see a home security blockchain interfacing with a home entertainment blockchain interfacing with a public blockchain as the Internet of Things develops. With a proliferation of blockchains the definite article in "the blockchain" would seem to have a very limited shelf life.

Blockchain: not the Messiah

Blockchain is an ingenious technological development, but it's far from being the saviour many are looking for

Wasteful by design

The most sinful thing about the current blockchain model (i.e. bitcoin) is that it is outrageously gluttonous when it comes to energy. It has been calculated that a single bitcoin transaction uses enough energy to power an average US house for a day (30kWh in case you were wondering). If these calculations are correct that is in no way sustainable. It's borderline criminally insane. "The energy required to run it would send us back to the dark ages," says Loosemore, expanding on the reasons why GDS rejected the technology.

This profligacy has not happened by accident, it's there by design, a feature of the "proof of work" used to randomise the verification of blocks of transactions by "miners". In order to prevent one miner or consortium of miners from monopolising the process (thus being able to manipulate the blockchain), verification is made intentionally arduous, i.e. processor and power hungry. Miners are essentially paid for their electricity plus a bit more in bitcoin to keep the system going. It is hard to see a way around this problem as proof of work is the central innovation of the blockchain.

The blockchain crypto-currency model is wasteful economically too. Because of the way it is set up a single bitcoin transaction costs an estimated $8.00 or £5.50, a burden that's currently being covered by investors, but for how long?

It takes a lot of time and space too. Used as a database a blockchain is currently only really suitable for storing and transacting very small amounts of information. "Blockchain bloat" is already a serious limiting factor when it comes to the scalability of crypto-currencies and it would be a lot worse if it were being used for moving data larger than crypto-coins. Worse still, blockchain can only manage three transactions per second (orders of magnitude less than even the slowest relational database), takes 10 minutes to commit those transactions and fails to acknowledge a successful write.

Stephens explains that for GBG, which is building its identity verification system on the bitcoin blockchain, the issues around writes are not much of a problem as the firm's use case is write-once-read-many.

"Three transactions per second is more than we need, but it is a problem if crypto-currencies are going to really take off. But there are really smart people working on how to scale out to many transactions per second. There are numerous techniques being talked about, but some of them remove part of the decentralisation aspect, and for us the decentralisation is key."

No doubt he is right, and many of these problems will be ironed out by the thousands of developers who are working on blockchain infrastructure and applications. For now though, blockchain is staggering under the weight of the expectations being heaped upon its slender shoulders. Blockchain is not the only mutual public ledger. There are other decentralised storage networks that offer far better anonymity and storage capacity without threatening to fry the world to a crisp, and there are few worse databases than blockchain. Which is not to say it is not impressive it is rather to point out that for every virtue there is a technological, legal or ethical vice.