Bitcoin has certainly made waves. Furthermore, its underlying blockchain technology appears to have widespread promise. However, it’s far from clear whether bitcoin has much of a future as a replacement for normal “fiat” currency.
Today, we’ll be hearing from Dan Hughes of eMunie – a man who argues that bitcoin is fundamentally flawed. What’s more, he thinks his own technology could ultimately replace it.
Does he really have a bitcoin killer? Read on, to discover his unique perspective on the cryptocurrency future – and then let us know your views: [email protected].
AL: What was the motivation for the eMunie project?
DH: Back in 2012 I was looking for a new challenge after discovering bitcoin. I was immediately convinced that this new technology would soon have a disruptive effect on the status quo. I decided to develop a service or product that could utilise it.
As I was researching, I realised that there were some technical and economic flaws. These could potentially cause some serious issues if bitcoin became successful. This prompted me to reconsider funding a project based on the bitcoin. I felt that these flaws might jeopardise the success of such a project – as well as squander my time and investment.
Instead I decided to investigate the possibility of solving one or more of these critical issues by developing a new platform. Thus, the eMunie project was born.
Initially it was just me – working solo on R&D. Over the past three years, the project has grown further – to include a small number of developers, a core advisory group, and a tight but expanding community.
AL: Can you highlight some of the issues with blockchain/bitcoin?
DH: All the concerns that we had can be categorised as follows: scalability, efficiency, price volatility, incentives and integration.
Solving just one of these problems in a satisfactory manner to enable bitcoin to progress further would take a substantial amount of time and effort. Of course solutions have been presented that address some of the above issues: improvements in hardware efficiency, volatility reduction, alternative incentives, and improved integration protocols have all been proposed.
In my opinion, all of the solutions presented are mere afterthoughts – late attempts to address issues that have been present since the beginning. The opportunity to implement preventative measures has long past.
The most complex and difficult issue to solve is scalability. There are no real, long-term solutions for this issue, without compromising the original aims of the project.
AL: Could you give a brief summary of the issues you’re concerned about?
DH: Scalability was the very first issue that we discovered – and is perhaps the most concerning. Blockchains simply don’t scale very well, hitting a ceiling at around 150 transactions per second. That’s about 7% of what VISA processes on an average day.There are technologies available now that can increase this. However, they are either inelegant – or involve some form of semi-centralisation. Such centralisation clashes with a core principle of blockchain.
The only option that wouldn’t compromise security, or require centralisation, is to increase the size limit of each block. This would allow more transactions to be stored in each. Currently each block is 1MB, and can store 2,000-3,000 transactions. A new block is produced every ten minutes.This equates to a maximum performance presently of roughly four transactions per second.Putting that in perspective, to be comparable in performance to VISA, each block would need to be around 500MB in size!
Transporting 500MB sized blocks around the network brings with it all manner of issues, such as connection saturation, propagation delays, etc. It also hits a tricky theorem called CAP head-on. This deals with data transmission issues, due to the finite speed of light. Simply put, the fundamental architecture of a blockchain is not well placed to handle transactional loads greater than a few 100 transactions per second.
Efficiency is another key issue, due to the manner in which the bitcoin network is secured.The primary method by which this is achieved is through a complex math puzzle referred to as proof-of-work (POW).Every ten minutes or so a new complex math puzzle contest is performed, much like an arms race among the computers on the network. They are working in competition, hoping to solve (or “mine”) the answer before anyone else does. The winner gets the bitcoin for that round of competition.Their answer to the puzzle is then checked by everyone and is established as the official “proof-of-work” for the transaction.
Having a substantial number of computers competing to solve the same math problem is a wasteful and inefficient method of operation. Currently, the bitcoin network is consuming around 10GW. Yes, that’s ten billion watts of electricity! This is required to secure the network, and the transactions which have been processed. Unfortunately, this situation will only get worse, as more users join in to try and solve the math problem.
Perhaps more critically, this approach doesn’t provide true immutability of data. An attacker with enough resources, patience and intent could potentially “rewrite history” across a large span of time.This would be a complete disaster for the entire system.
Price volatility is probably the most obvious feature, to anyone who takes an interest in this technology. Historically, digital currencies have been extremely volatile. While this is due to a number of issues, the root cause of such volatility is the implementation of a fixed supply model. If supply is forcibly limited, the only variable left to adjust is the price. All digital currencies to date have used this method, and it results in a much more volatile price than would be the case if supply could also change in response to demand – just like it does in ordinary markets. If cabbages are selling well, more farmers will grow them – increasing supply.
Bitcoin mining has rapidly centralised, due to the ever increasing barrier to entry. Nowadays, only specialised computers have a chance at solving the math problem. Miners need to carefully manage the return on investment of equipment they have purchased to engage in the mining arms race.
Another issue with the incentives policy lies with the “full nodes”.These are computers in the network that just want to do basic transaction processing and provide a copy of the blockchain. They don’t participate in the expensive POW operations. Nevertheless, they store a copy of all the work and transactions ever done since bitcoin began, and they deliver that data to new nodes joining the network. These actors in the system receive no rewards or incentives for their sacrifice of storage space (currently ~100GB), or for their electricity usage. Over time, the number of computers with a complete copy of the network may reduce – as these users may no longer have the space required, or they see no financial reason to do so.
Finally, I had a concern regarding the lack of integration with existing infrastructure. Almost all current infrastructure integration is achieved via third-party centralised services – such as exchanges, debit cards and online payments. This effectively defeats the purpose of the project. Furthermore, all of these services apply a fee, in order to support the service. In some cases, this can be as much as a few per cent.
AL: How does your project, eMunie, propose to solve these issues?
DH: Initially, we looked at solutions we could apply to the existing technologies, but it quickly became obvious that this was not viable.
Instead, we decided to go back to the drawing board, and look for alternative solutions. We needed to solve the aforementioned issues – and yet remain true to the ideals of bitcoin and blockchain technology.
Almost all of the technology in eMunie is new and developed from scratch. We started with the scalability issues – and decided that the only solution was to develop a new form of ledger, which took into account parallelism.
AL: Can you explain what you mean by that?
DH: To increase performance in a blockchain, the only real option is to add processing power to each node. This may not be viable, and is ultimately finite. Our approach partitions the work into more manageable chunks, so that nodes in the network are collaborating together to process transactions. Collectively, they achieve greater scalability as a whole.
Furthermore the architecture of the ledger is such that it supports a number of services natively. This allows easier access and utilisation of existing infrastructure without the need for third-party services (middlemen). One such use case is the ability for a user to create their own legitimate debit cards using ~£20 of easily purchased hardware. This removes banks and card issuers from the process. Despite this, users can then purchase goods from a bricks-and-mortar store a few minutes later.
Next on the list was the consensus algorithm (ie, math problem) and we wanted to develop a solution which was comparable in terms of security to the existing proof-of-work, yet much more efficient, with a lower barrier to entry.
To achieve this we took all of the positive functionality of existing consensus algorithms and investigated how we could use one or more of them in a manner that would achieve our goals. The result is a consensus algorithm that utilises components from a number of current solutions, which are then modified in such a way to achieve similar security – but with much improved efficiency.
Finally we tackled the problem of volatility and took some inspiration from how the current banking system operates. Instead of a fixed supply issued over time, we decided to implement an elastic supply currency that would self-regulate to demand in real time, issuing and removing currency from circulation as required and stabilising the price.
Traditionally in existing banking, this is a tricky balancing act as there are a number of factors in play that exaggerate the difficulty of achieving currency stability. These factors include human error, and delays in obtaining economic information.
Removing humans from this process removes an obvious source of error. Furthermore, automation means that the network is able to react quickly to any economic situation.
AL: How can this model be tested?
DH: We wanted to be able to test using real-world trade data of a volatile currency or commodity. As the project is still in development, we didn’t have any native trading data, and we didn’t want to resort to simulations.
Instead, we decided to collect as much trade data for bitcoin as we could. We executed all of those trades through our economic model, to see if it was possible to stabilise bitcoin to some degree.
Our first test consisted of 500 million-plus trades spanning a number of currencies (USD, GBP, yuan, EUR) and the results were simply unbelievable. We assumed that we had made an error in the parameters we were using – so we performed the same test a number of times, with varying sets of parameters to test all possible cases.
The result of all these tests were broadly similar. We were able to stabilise the bitcoin price using our economic model to <1% per annum of gross deviation.
AL: How does a stable value affect potential adoption?
DH: This is a question I am asked regularly, specifically concerning initial adopters who are most likely speculators. There is generally confusion surrounding how these early adopters would see any ROI [return on investment] and thus an incentive to invest.
Before I touch on that, a more stable value brings some obvious and immediate benefits – particularly when it comes to trade between consumers, merchants and businesses. One of the primary reasons that we feel bitcoin and cryptocurrencies in general have not been embraced by the masses so far is due to the gross volatility in price of each unit.
That volatility puts additional burden on the users, as the only way to be sure that what they have today is worth the same tomorrow is to either spend it, or trade it immediately for a more stable currency (usually USD). If the consumer doesn’t spend it all, or the merchant doesn’t cash out their takings, then they run the risk of their holding losing value.
There are services such as BitPay which smooth out this process on behalf of the merchant. However, with third-party services come third-party fees – which ultimately places the cost of accepting bitcoin in the same range as accepting regular fiat money. That completely removes one of the main purported merchant benefits: cost.
With a more stable value, merchants will be more inclined to accept it as they are not held to ransom by the volatility, and are therefore no longer forced to sell immediately. The same goes for users – as they will be happy to hold it for longer periods without worrying about their currency devaluing before they have spent it. More users equals more merchants, and vice versa.
Early adopters need incentives to participate – and these are present, just as they are with traditional fixed-supply cryptocurrencies.
I’ve actually been quite surprised how much focus is put on price – to the point where the quantity a party holds is completed overlooked in the speculation process. With the eMunie elastic supply, it is quantity that provides any ROI, and not price.
To take a simplified example: assume that eMunie is in growth. The economics model will be monitoring demand vs supply for any deficit. Upon it detecting a supply deficit, it will create and distribute new supply fairly around the system – with payouts to those who have a balance, and to those who have provided computing resources to ensure operation of the network (we’ll ignore the latter for the purpose of this example).
Let’s now assume that there is a 100% deficit in supply. The system needs double the amount of EMU [eMunie’s currency] to meet demand and so creates it – and distributes it.
In this scenario, if I was to hold 100 EMU, I would receive an additional 100 EMU and the value of my holding doubles as the price of each EMU is still the same. The price of something increasing from $1 to $2 is exactly the same ROI as the quantity I hold increasing from 1 to 2 – but the ROI is simply delivered in a different, perhaps less intuitive way.
The best way to think about returns on deposits is as “interest on savings” in a bank. Periodically, you receive a payment of X USD/GBP/etc – and that is the ROI on your savings. The same principle is employed with eMunie.
I’d just like to point out that reducing the supply in a demand deficit is a different process and doesn’t “haircut” user balances to achieve it. More information on that can be found online.
AL: Can eMunie technology be used in private industry?
DH: Absolutely, private industry has been part of our focus since the project began. We wanted to allow easy creation of private networks, and also to enable these networks to bring value to each other. Just as with the bitcoin blockchain, the eMunie technology can be repurposed.
As a ledger is simply a record of events over time. Any industry that has this requirement can benefit greatly from this technology. There are many applications, which offer improved efficiency, whilst reducing costs.
For example, imagine a private healthcare network, used to store patient data and reduce administrative costs. A user in the public eMunie network can download an app to his device, which can communicate securely with this private health network.
He can contact his doctor, request repeat prescriptions, and view his health records. This could be very useful in emergencies – especially if the patient is abroad. Currently, sending records can take 24 hours or more.
AL: What are the remaining tasks before launch?
DH: We’re now into the final stages of development. We hope to have the majority of the critical components finished within the next few months.
The platform will undergo heavy testing, to discover any possible issues. We need to ensure the security and stability of the platform. Additionally, a full external security audit will be performed before we move to a final release.
Other tasks include completion of the technical documentation; development of mobile/tablet clients; and EFTPOS software development for integration of our payment protocol into existing merchant solutions.
AL: Thanks for your time, Dan – and good luck with the project.
DH: Thanks.
Best,
Andrew Lockley
Exponential Investor
Category: Investing in Bitcoin