This is the second part of our exclusive interview with Hyperledger’s executive director Brian Behlendorf. You can read the first part here.
The explosion in interest in distributed ledger technology over the last few years has created a number of unique problems for players in the space. A technology that is focused on decentralization of the mechanisms of central control of information throws up a number of interesting governance issues. How does the community make decisions that will impact the things developers can do and the experience of end users? How should disagreements over direction and strategy be handled?
Another interesting problem comes from the continued, and increasing, interest of central banks, governments and international governance organisations in the technology. What effects will these centralized authorities have on a technology that is primarily thought as stripping power from the centre?
These are all questions that the members of Hyperledger have to think about as they try to drive the technology forward. As we saw in the first part of this interview, while the organisation is somewhat insulated from the intense speculative activity that is currently characterizing the blockchain and wider cryptocurrency space, it still needs to develop technological solutions that serve the commercial needs of its members and end users.
Any organisation that is looking to revolutionise marketplaces or other information networks that today require central market makers needs to dedicate a lot of energy to thinking about what governance structures should replace the existing models.
The Block sat down with Brian Behlendorf, executive director of Hyperledger, to talk about governance, security, the fundamental importance of the ‘right to fork’ and how to ensure governments and other centralized entities don’t compromise the essential promise of blockchain technology.
The quest to re-decentralise
Collaborative ways of working and the decentralisation of power and knowledge has continually appealed to Behlendorf. In 1991, when he first got a grip on the fledgling Internet, it was fairly decentralised but still able to work as a coherent system. Since then, he has watched as the net lost that dispersed aspect, and he couldn’t help feeling that we had lost something: not just aesthetically, but potentially economically as well.
But that began to change in the latter part of the last decade.
“I had seen plenty of technology efforts to look at ‘re-decentralising’ the web, and started to pay more attention to them,” he recalls. “Then the abstract of Satoshi’s paper when it came out in 2008, and I started to kind of follow the Bitcoin space but never invested myself, much to my chagrin!
“The idea of currency speculation, it’s just not innate to me and a lot of people, the idea of currency as a basis for how we reinvent systems. Coming from a world where you didn’t set up sites to monetise every transaction, you didn’t set up email servers to charge people to send you email, you set up your own servers to collaborate.”
The emergence of a collaborative community of developers around cryptocurrencies and its underlying blockchain technology showed Behlendorf that he was not alone in his interest in re-decentralisation of the internet. He began keeping an ever-closer eye on what was developing in the space.
“Things started to heat up, and it was when I was working as a partner in a venture capital firm,” he says. “We started seeing tons of these bitcoin and blockchain companies, which only made me more sceptical, until I saw one talking about land titles in emerging markets.
“That seemed like such a small market opportunity, so shouldn’t have been that interesting, but as I spent more time understanding it I saw it wasn’t the market size that intrigued me but the prospect of solving this entrenched problem – more digitisation but society leading to more centralisation.”
The central question here was whether a cryptocurrency was actually needed to make distributed ledger technology for land titles, and other industries. The answer seemed to be a definitive ‘no’. It did seem possible to implement the technology as permissioned networks, which would answer a lot of tricky issues – such as what happens if there was a theft in the system or if an individual did something contrary to the intentions of other market participants. On one end of the spectrum, in the world of Bitcoin and Ethereum the idea is that there can’t be any reversals. There was, however, a reversal when a problem turns to be serious enough, such as when DAO suffered a hack last year.
“Basically, the human governance kicked in and said ‘we’re going to correct’ – which the technology allowed – but we think it’s a bad idea,” says Behlendorf.
“The community has grown so much larger, so much more fragmented”
Security and governance
But when he first sat in his new office and thought about the first steps he should take, it was not marketing or releases that were on Behlendorf’s mind. He had something more fundamental to take care of first:
“The first hire I made out of my own budget for full-time work on Hyperledger was a security man whose name was Dave Huseby. He came from Mozilla – he helped maintain Mozilla’s Tor patches for a while – and was very well connected into the crypto-community and very sharp with regards to how this is all supposed to work,” he recalls.
“I didn’t hire him to scrutinise every line of code, I hired him to institute a culture of secure programming that help us prove to the world that we are prioritising the secure development of these applications.”
Part of that culture is maintaining a close relationship with higher security organisations that can do regular security scans of the code, but it also means putting in what Behlendorf describes as a ‘critical bug-binding process’. The aim is to create a system that incentivises and rewards security researchers who find holes in the code or other processes. While this all shows how fundamental security is to the project, and how seriously it is taken, the problem space is very different.
“Because it’s not an anonymous network, because every transaction is, every smart contract, everything done on the network has a human attached to it. There is an appeals process and it may not be that everyone gets refunded their full money,” says Behlendorf. “I mean, that is going to be up to each specific deployment to decide what their governance model and appeals process.
“I think there will be something like what happened on the Swift network, you know, where someone hacked the Bank of Bangladesh and laundered $87 million to a casino in the Philippines. It is a major problem and the question of who is liable for coming up with the missing cash is not for the technology to answer, its beyond us to answer.”
For Behlendorf, the answer to this question from Hyperledger’s perspective is to say that if the system recorded the transaction and where it came from, then it is up to Swift and their networks and banks to decide how to proceed. That is, after all, what institutions like insurance are for. When it comes to blockchain, the problem is different. There is no recourse to get back the $30 million stolen from Parity wallets without undergoing a massive hard fork, like the one that DOA went through in 2016.
Behlendorf is sceptical that such a hard fork is even available to the community anymore:
“The community has grown so much larger, so much more fragmented, and like we see with the Bitcoin community where the interests of some parties can diverge drastically from the interests of other large parties on that network. This can lead to very vicious debates, and worse, a currency split.”
For people who believe that the trust in a system comes from making these kinds of forks hard to do, the type of hard corrective action DOA underwent can be seen as a compromise or a failure of governance. For Behlendorf, it represents an important part of the process of trying to figure out what kind of systems people want, and what kind of technology those systems will need.
“There are lots of thorny issues there and as technologists we should be building tools that support whatever governance model people want to come up with, frankly, and if they want a very rigid one or a very permissive one. There should really just be a dial.”
“I have always seen that in open source software, what we call the ‘right to fork’, as a strength”
The ‘right to fork’
Differences in opinion exist over the course of blockchain and cryptocurrency governance – and their potential effects are a hot topic in the space at the moment.
The reason: the recent hard fork of Bitcoin, which resulted in the digital currency splitting in two. The fork occurred in large part to disagreements between Bitcoin’s core developers and miners over whether the size of Bitcoin blocks should be increased.
While this hard fork has been seen by some as a failure of governance, for Behlendorf it represents one of the benefits of this kind of collaborative work. “I have always seen that in open source software, what we call the ‘right to fork’, as a strength,” he says.
“This right to fork is an essential check on tyranny and I think it is better off open source because it means that you can rescue projects that are under bad leadership, and now we can transpose that to financial networks or to other times of information networks where the traditional, central big data hub model is an up-down centralised model.”
In a decentralised data system where the goal is to create some kind of central market utility like Swift, then the system begins to act like a private market regulator. It does this by setting the ground rules that people follow in order to use the technology stack. For the system to work in a functional way, there is the need for standardised talk and effective appeals processes to be put in place. Behlendorf believes that the ability to fork is an essential part of the process of making blockchain markets as effective, democratic and competitive as possible:
“Well, if everyone has a full copy of the ledger and they decide collectively that the organisation that they’ve bestowed as the permissions gateway for who can get on that network is no longer serving their needs, they have the ability to change,” he says. “Much more easily than if that central market actor actually held and controlled all the data, and controlled the world’s data.
“That will lead to more democratic markets, it will lead to more competitive markets, and while it might not achieve the crypto-anarchist vision of no regulators and no governments, it will probably make them smaller and more competitive and more focused by making it harder for them to own and control.”
“part of the escape valve here and the reason I sleep pretty well at night is that we are not talking about one blockchain”
Here come the centralisers
As blockchain technology has continued to gain momentum with enterprises across a range of industries, it was perhaps inevitable that governments and other institutions would begin to take notice.
The last year has seen the United Nations setting up a blockchain working group, Dubai proclaiming its aim to be the first ‘blockchain-powered government’, and a whole host of central banks announcing that they are testing the technology.
Does the introduction of governments, who are be their very nature centralising entities, clash at all with the desire to decentralise that sits at the heart of the technology’s development?
“I take a fundamentally positive view for exactly the same reason that as governments started putting up websites,” says Behlendorf. “ That seemed like a pretty positive development, right? Part of the escape valve here and the reason I sleep pretty well at night is that we are not talking about one blockchain.”
For Behlendorf, the arguments over whether government involvement begins to damage the integrity of, say, the Bitcoin ledger when they try to use it track criminals, become mute when talking about a world made up of a multitude of different blockchains. It is likely that the future will see different blockchains servicing different tasks for different markets and across diverse geographies.
“I mentioned land titles as one of the use cases that pulled me in,” he continues, “well there is active discussion now on whether the land title chain sits by itself, or do you also use it to record mortgages and other derivatives products? And then, is it the size of a local county, which in the US all of our land title records are actually maintained by recorders offices at the county wide rather than the city or state level.
“But does that make sense in a blockchain context, or should we be talking about systems that are state-wide or even US-wide for tracking land title. Or even larger?”
While it does seem true that there is more trust in larger networks with more participants, the constant possibility of setting up new chains and forking will likely result in there being as many blockchains as there are governance models. A blockchain for managing property and one for major payments that look distinct in terms of the code, consensus mechanisms, smart contract language and other technical factors.
“If you want to buy a house using cryptocurrency or digital assets that are correlated to shares of bars of gold, then you will do the transaction across two chains and your done with it. That will be the interesting thing to watch, and that’s why I am not worried about governments stepping in.”
What does the short-term future hold?
So, in the middle of a space that has as many eyes on it as blockchain does, what does the immediate future hold for Hyperledger? Behlendorf wants to stay focused on the organisation’s primary goal and not get swept up in external events:
“There is so much activity, which is great, but also a lot of noise in the space so it’s a lot of work to stay on top of the fray.”
The near future is essentially business as usual. “My goal is both to recruit companies to be members so that we have the resources to go out and build the community, but my real metric is developers. How many people are actually involved in building the code? Are we everywhere we should be? Do people think to turn to Hyperledger Fabric or Hyperledger Sawtooth or our other technologies as they’re building new things?” While there is not likely to be any serious M&A activity, 2017 could still see another release along the lines of Fabric.
“The nice thing about being involved in this project is that everyday someone shows up with something new and we’re very opportunistic,” says Behlendorf, smiling. “So, we’ll see what happens.”
Find out more about Hyperledger’s work here.