School of Hard Blocks
Employeeless companies and blockchain squabbles… Chris Thornett meets Dr Catherine Mulligan to talk blockchains and digital economies.
Employee-free companies and blockchain squabbles… Chris Thornett meets
Dr Catherine Mulligan to talk blockchains and digital economies.
a new mindset is required “Often digital technologies are built without citizen engagement – it’s done by an industry.”
Dr Catherine Mulligan is one of the very few people in the world who’s both an expert in telcos and blockchain, which she admits makes her “quite weird.” Actually, she’s quite an extraordinary person as, among many things, she’s a visiting research fellow for the Innovation and Entrepreneurship group at Imperial College London, fellow and expert of the
World Economic Forum for Blockchain Technologies, and while we were arranging our interview she was at the UN, sitting on the Secretary General’s high-level panel on digital co-operation. Dr Mulligan was at the Open Networking Summit in Amsterdam in October 2018 to outline the possible future use cases of blockchain technology, although Arpit Joshipura, general manager of networking at the Linux Foundation, was clear to indicate that blockchain is “three to five years away” on the Foundation’s networking roadmap. Dr Mulligan also spoke about DataNet, a UCL project she’s leading that aims to develop internet-like robustness and standards for data by developing unified protocols and a global, trusted infrastructure.
Linux Format: What does digital co-operation involve for the UN?
Dr Catherine Mulligan: I had my first face-to-face meeting in New York recently. During the general assembly, apparently, the secretarygeneral got up and said this is his second highest priority after climate change, the panel that I’m sitting on. The first question we’re all asking is, you know, what does digital cooperation
mean in this context? But I think it’s ways to understand what it means to engage a much broader set of people in the process around digitalisation and that touches on a lot of issues. So you’ve got, for example, the inclusion of different countries; inclusion of different people; the issues around data privacy. The fact that quite often digital technologies are defined and built without citizen engagement – it’s done by an industry. Obviously there’s a lot of different aspects involved and the panel’s website will be kept updated with information as it progresses.
LXF: On the remit for the panel?
CM: Yes, on the remit. As with all these things, there’s a lot of discussions that sort of have to happen. But I’m there contributing things around digitisation, digital transformation and blockchain, obviously. Blockchain is what I’m kind of known for, but I was doing a lot around digital transformation previously and a lot around, for example, IoT, smart cities… those kinds of things.
LXF: Is this anything to do with the ID 2020 initiative ( https://id2020.org)? CM: No, it’s a separate panel, but the idea is that we’re going to do a large amount of consultation with other types of people who are doing digital coordinations. We’ve talked to the broadband forum, GSMA – or the people on the panel have – there’s the Linux Foundation obviously, which is another way to perform digital coordination and cooperation.
And then, you know, I’ve done previous work around something called open agile smart cities, which was taking citizen engagement – bottom-up, grassroots citizen engagement. These were city leaders, academics, startups and accelerators, the European Commission and other government bodies and also large-scale industry, and then putting them together and through that sort of process defining how you will take something into ITC [International Trade Centre] or something into ITU [International Telecommunication Union]. The idea is that you get money and more voices heard in those forums and then those submissions, rather than just what has so far been pretty much large industry or government only.
LXF: That’s one of the worrying things about blockchain, the people who code the blockchain have quite a lot of power. CM: It depends what you mean by the concentration of power. I think we have a lot of concentration of power already in industrial structures. Take Android: we did an analysis one year of seeing how many vendors actually got their code accepted into the core part of the Android platform. Approximately five. You know – despite the hundreds of people attempting to contribute! So the concentration of power always existed and what’s interesting about blockchain is the way that’s manifesting itself. I mean, people are people and you do have an issue.
So with a lot of blockchain solutions, for example, I walk into rooms and it’s always young white men. This could be extremely dangerous if we think about the fact that what many people are positing is that blockchain can perform services that were previously operated by government or parts of industry or local authorities or things like that. That means then, how do we ensure that everyone’s voices are heard because that’s effectively enacting a new form of social contract in the world.
LXF: Does that mean we need to apply some ethical training or something along those lines for coders that are coming up? CM: Yes, I think there are two things, so yes to ethics. I think it’s something that’s required in civil engineering. They talk about ethical training as part of what they do and it’s something that I’ve been pushing for with anyone who will listen when I talk about computer science education [laughs], which is that, we need to get people to think about ethics and about the social contract about the impact of what they’re doing.
The other thing that I think we need and if we take a corollary from other industries – law, accountancy, civil engineering – all of these sort of industries where previously people have realised they have a huge impact on society, you need an accreditation. So you’ll do your undergraduate and then you have to work professionally under supervision, before you go through a certain number of postgraduate training certificates, right? And that means that you’re than qualified to practise on your own.
I think that we’re getting to this point these days with software. If we’re going to create smart cities that are controlled and operated by some hardware but mostly software, then we need to educate and potentially certify our computer scientists
in a new way.
It’s no longer just about technical brilliance, but the problem is you can’t just ignore the technical brilliance either; you’ve got to make sure you help people become both. I mean, if you look at, for example, the DAO hack, which was a big hack a couple of years ago in the blockchain area – that was built on Etherium. If you look at what that was and it was two lines of code, which were basically in the wrong order. The fact that it was a Turing complete language in the cloud, which meant that somebody could exploit that repeatedly in a loop to take out, I think, it was about $50 million.
Think about people who are saying cryptocurrency is going to redefine financial services. We need to ensure that the code is safe enough to us for that because currently, we have huge amounts of regulation that protect consumers in the banking system. In the UK you have £75,000 protection, that’s the government saying the £75,000 is protected. You don’t have that with cryptocurrencies.
And the other thing is that everyone says blockchain is more secure. Yet while nothing in blockchain itself has been hacked, everything in the ecosystem around it has: wallets have been hacked and exchanges have been hacked.
LXF: Am I right in thinking that the permission blockchains can avoid some of the speed and capacity issues?
CM: Absolutely. I didn’t mention it during the presentation, but there’s this difference between permission-less, shared systems and permissioned private shared systems. If you remember the spectrum [in the presentation] from permissioned all the way through to the blockchain bitcoin ones. As you move across that you get greater decentralisation, but you also get a reduction in transaction speed. The number of transactions you can do reduces quite significantly and that’s because of a couple of ways that the Bitcoin network is structured or the Etherium network is structured specifically to prevent double spend.
But with the enterprise level ones, like Hyperledger [ www.hyperledger.org], you’re able to get a pretty good transaction speed through. It’s something that when we talk to people about blockchain, it’s like one of the core pieces that you tend to ask:
What sort of transaction speed do you need? So for many of the world’s transactions the throughput that you get on the new versions of Fabric [ www.hyperledger.org/projects/fabric] is probably okay and then some of them, for example, financial services you really need higher speed transactions unless you’re talking about remittances, in which case you’re sort of talking about a month usually to do it. In which case the speed is less of an issue and it’s more about KYC...
LXF: KYC?
CM: Know Your Customer. There are antimoney laundering (AML) laws, too.
LXF: Sorry, I’ve spent the last two days having loads of acronyms thrown at me and waking up in a cold sweat screaming I know what an SDN is, now let me sleep! CM: [Laughs] What’s an SDN. What’s an SDN-C, for that matter!? LXF: You’ve mentioned distributed autonomous organisations (DAOs). What are they and what interesting examples have you encountered?
CM: As I said, I’m a computer scientist and an economist for my sins. It’s a fascinating way to think about the future and the impact of digital on society or economy. The idea behind a DAO is that you’d have companies which have no people in them. That sounds counter intuitive, but actually, some people have built very small versions of this that are running, for example, websites. So the DAO can go out and purchase, say, more bandwidth. It’ll just create an electronic contract and go out and buy itself more bandwidth. If it needs service, then it will send a service request, you know, and actually, a lot of these things now are automated.
talking decentralised platforms “You could quickly imagine an autonomous vehicle version of Uber, where we get rid of the drivers. Then you can think about one that’s decentralised…”
A bigger example that would have an impact on humans would be if you think about Uber and the centralisation of those platforms. You could quickly imagine an autonomous vehicle version of Uber, where we get rid of the drivers. Then you can think about one that’s decentralised.
So let’s say you and I own cars, as do people in our town. Ninety per cent of the time they’re unused, so the theory would be that using a decentralised platform you’d have a company that has no people in it, but decides when and where there’s spare capacity in your vehicle and rents it out to somebody. Then if I get into the car, I would then pay with a cryptocurrency.
There’s a lot of areas where you could see such things being applied. For example, in the IoT areas, you could imagine machine economies being built because one of the biggest problems with IoT has always been this idea of monetising it. The problem is the transactions for IoT devices are so small that there’s not enough money in it for a human-run organisation. So you can see machine-run companies operating autonomously. They don’t need to rent space; they don’t need to have an office building, food, paper…
LXF: Salaries…
CM: Exactly! So you could imagine an IoT-
enabled marketplace that’s all DAOs, all interacting with each other – buying and selling stuff that way.
LXF: Wow. Yes, that’s a good one, I like that. We’ll see if we can illustrate it.
CM: [Laughs] There are actually some good videos that you can probably look at for inspiration of how a DAO would work
LXF: Time-lapsed branches being sprouted and all that sort of stuff?
CM: Exactly. I think the real issue for blockchain is how do you explain this to people? Once you get it, you get it, but I’ve been working with it now since about 2010. So, you know, I’ve had eight years thinking about these things and it’s not like when you come to the ONS and people are talking about we’re going to virtualise the networks and we’re going to put stuff on to cloud, which I think is quite intuitive for most people who’ve worked in networking before. But it’s not so intuitive to say: well, you’re going to centralise everything and you’re going to remove a lot of the processes out of an operator network and you’re going to rely on machines to do it. I think that it’s a bigger leap for people.
LXF: Looking at the issue of governance for blockchain. From a personal perspective, I was intrigued to hear about Civil for journalists [https:// civil.co] for newsrooms and using a blockchain system to vet content, but that kind of thing brings up the issue of governance. Is that something that needs to be dealt with for a lot with these new technologies coming out?
CM: Yes, so it’s not just a problem for blockchains, but for a lot of technologies. And I think it’s a problem for the world to solve. You know, how do we solve it for privacy? How do we solve vetting of information sources that gets put onto Facebook and stuff?
With regards to blockchain however, I think it’s something that gets critical, because it’s exactly this idea of a decentralised system. If you’re using Bitcoin and Etherium, you need to establish what the governance for that particular network is going to be, because once it’s up and running it’s difficult to change. And we’ve seen that, for example, Bitcoin, people arguing about what the right block size was going to be a couple of years ago and it was a beautiful example because you could achieve consensus on the protocol on a technical level.
Yet as soon as human beings got involved in the process, it spiralled out of control. People very publicly fell out with one another who had been friends for 20- odd years and writing nasty blog posts to each and all on this. So it was a wonderful example of technology you can’t rely on to solve all of your governance problems.
Then for the enterprise-level blockchains, like Hyperledger, you need to work out the governance before you implement them. And you have to implement the governance procedures before you even build the technology. That’s because you’re fundamentally changing how a business process works and it means a lot more discussion with lawyers and things like that before it all comes together. So it’s a critical issue.
LXF: Can you tell us about your DataNet project you’re working on?
CM: We’re looking at providing the same level and robustness of infrastructure for data that we have for the internet, which has been structured into three layers.
On the bottom layer you’ve got physical transport standards, which are done by ITU, and you’ve got that logical layer, which is like ICANN and the IP sort of connectivity layer. Then you’ve got the layer on top that’s emerged recently, which is basically data in the economy and society, right?
So it’s moving outside of corporate boundaries. And the problem is that it hasn’t been organised particularly effectively, which is how we’ve got massive manipulation of data sources that are from privacy invasion, through to not being able to trust new sources that come through, for example.
The idea behind DataNet is to provide the same level of trust and identity at the data layer that we have in the internet layer, which we trust to a relative extent at the moment. We’re looking out quite far in the future, but the idea is that everybody would be able to manage and control their identity and everyone would be able to manage and control who has access to their data, and how that data is used. On the other end of that spectrum, if you’re a company then the company knows this is the correct date for this particular person and I can trust those data sources if it’s coming from an IoT device or something like that. The idea is to create the sort of trust and locatability in data that we currently have in the internet.
LXF: But controlled by the individual rather than by some corporation, or would that be outside of government as well or not, or within some sort of government structures?
CM: The government needs to assess where regulation needs to be set. A lot of people think it shouldn’t be regulated at all and we should just leave it open. And so that’s a personal perspective. If you take Estonia, they’ve done a brilliant job of rolling out infrastructure to the entire country. Some of that is based on blockchain adjacent technology – or so we call it. It’s not blockchain 100 per cent, but close.
So there are two things here. One, they’re able to trust and they’ve created economies on top of that – new economic structures that sit on top of their digital infrastructure. And while I don’t necessarily think that the rest of the world should replicate exactly the same example, I do think it indicates there’s a need for some kind of data rails as a national infrastructure. We should be treating data and communications infrastructure in the same way that we treat water or electricity. So everyone has rights to water and electricity…
The market structure around that is something government needs to set, but we need a national infrastructure of some sort that provides data integrity, data security and data privacy.
LXF: Data is the new bacon. CM: The new bacon, yes.