Growing with Blockchain

From disruptive potential to operational reality

Kevin R Smith

• Publisher: Novaro Publishing
• Year: 2020
• ISBN:
  9781838067410
• Topic: Negocios y empresa
• Subtopic: Comercio electrónico

Following lockdown, the digital trust that blockchain creates for networks of remote users started to re-define relationships between customers, suppliers, partners and funders in earnest. As a technology, its ingenuity is already widely recognised. The question now is who can best deploy it in the market to raise money, make payments, negotiate deals, engage customers and manage suppliers. This title draws on the knowledge and experience of 18 top-level blockchain performers, including IBM, Dell and CMS, to bring a wide audience up to speed with the options and actions for building and growing a business with blockchain.

1.
BLOCKCHAIN AND THE INVENTION OF DIGITAL TRUST

Robert Learney at Digital Catapult reports on why blockchain represents such a radical development in computer science and on how the trust machines it creates are being deployed

‘I’ve developed a new open source P2P e-cash system called bitcoin. It’s completely decentralized with no central server or trusted parties, because everything is based on crypto proof instead of trust. Give it a try or take a look at the screenshots and design paper.’

It was with these words that Satoshi Nakamoto revealed the invention of blockchain to the world on 11 February 2009. As you’ll hear from many different sources, the blockchain created the first way for people to exchange or transfer value over the internet without a trusted third party. It’s called ‘the trust machine’ for this very reason: before blockchain, no trust without third parties; after blockchain, suddenly a way that both you and I alone can agree that what we see written in front of us is the incorruptible truth and that, yes, I did send you those five bitcoins, bananas, invoices or whatever.

But Satoshi didn’t work in a vacuum. His (her? their?) breakthrough actually required nearly 70 years of groundwork, combining the wisdom of hundreds of computer scientists and advanced mathematicians from nearly every corner of the world. It’s this journey from there to here that makes blockchain even more revolutionary and sets the scene for any discussions of its future.

The trust machine

So what exactly makes blockchain a trust machine? Without repeating lengthy explanations of hash functions and elliptic curve cryptography, the basics are as follows. You and I both agree to use a notice board to post messages. But if I post a message to tell you I’ve sent you five apples and also try to post one telling Mrs Smith that I’ve sent her three of those same apples, the notice board sees me trying to lie and stops me. In fact, it all depends on which message it accepts first: the five apples to you or the three to Mrs Smith. In either event, I can’t try to lie to the notice board, which means I can’t lie to you (or to Mrs Smith).

So how is this different from us all agreeing to use something like a spreadsheet stored in a cloud database, where we each have the right levels of permission to read and write? It’s a perceptive question. In many ways, it would look and feel much the same to you as a user. But in some deeply fundamental ways, a blockchain is completely different. With a blockchain you don’t need to trust the company running the cloud storage that they will keep the system up and maintain its audit trail or that they hadn’t made some secret changes to the software so that Mrs Smith always gets one extra apple.

In fact, while cloud databases might not explicitly use blockchain, they do make use of some of the same concepts that led to the creation of blockchain.

From the beginning

The first computers built in the 1940s were isolated, standalone systems consuming the power of a whole house. Invented for war, these first machines were set to tasks such as cracking ciphers and calculating artillery trajectories. It wasn’t until 1951 that computers were used for civilian purposes (for tabulating the 1950 US census).

The 1950s saw developments in miniaturization, storage media and initial uptake of computing machines by the business community, but most importantly for the history of blockchain, the invention of the modem by the US Air Force in 1953. Computers could now communicate over normal voice lines rather than telegraph or special purpose data lines.

Making use of modems, telephone lines and teleprinters, the US military embarked on a project called Sage (semi-automated ground equipment) that used a series of massive computer centres (in fact, the largest physical computers ever built) to co-ordinate military radar data from across the US into a unified picture of US airspace to defend against Soviet nuclear strikes. From 1958 to 1966, 22 Sage centres were built and they remained operational – vacuum tube computers and all – until the 1980s.

Things then progressed rapidly throughout the 1960s with the creation of new programming languages, the invention and first use of integrated circuits, and new storage media. But most importantly for us, networking developed to the point of allowing people to log in to large central computers using remote terminals in a hub-and-spoke model. Note that the computers themselves weren’t yet networked, but users could post messages to each other using the precursors of email and bulletin boards.

True inter-networking had to wait until October 1969 with the activation of the ARPAnet, the forerunner to the internet. For the first time, tens of computers across the continental US could share messages over a common network using a standardized protocol. This grew to hundreds of computers and soon reached beyond the US into Europe via transatlantic links. The rest, as they say, is history.

But there is one more important parallel invention which opened a new line of thinking which we need to explore first. The database. It may seem trivial. After all, people have been storing data for thousands of years in many forms. But the creation of digital databases didn’t take place until the 1960s. Interestingly, one of the earliest databases evolved directly from the US military’s Sage programme to serve bookings and reservations for American Airlines. Installed in 1964, the Sabre system (semi-automated business research environment) used two IBM mainframe computers and could handle 84,000 queries per day. The 1970s through to the 1980s then saw the development of more recognisable relational database systems such as SQL.

The birth of distributed computing

The combination of databases and networking led to the invention of distributed computing as its own branch of computer science in the 1970s. Distributed computing is exactly what it says: a method of dividing up or co-ordinating computational tasks between multiple physically separated systems or nodes.

When you think about reconciling databases, it is exactly what you need to do. You and I both need to make sure that the database telling me how many apples I have matches your copy exactly. The data we’re sharing is known as the state of our system. Our computers have to communicate with each other to update the numbers of apples on our databases (state) and we have to make sure that neither one of us can let our copies fall out of synchronization with each other or one of us might try to sell the same apples twice. This is the nature of consensus: our two systems are marching in lockstep with each other.

At first it might seem obvious that you should connect multiple computers in a way that has one master computer co-ordinating the activities of a large collection of subsidiary systems to update their databases and manage computations. This is the classical client-server architecture, but it will fail if the master computer misbehaves. It could be misbehaving on purpose due to malicious code or it could be because a cosmic ray flipped a bit in memory and now the system is off running in its own little world. Either way, you have to trust absolutely in the stability and availability of the master.

So instead, let’s make every node holding a replica of our database capable of behaving as both a client and a server, a so-called peer-to-peer architecture. We’ll get a much more resilient system as a whole, because now any single node can misbehave while the rest should be able to carry on as normal.

The byzantine grandparents of blockchain

But the problem was that nobody at the time knew how to build peer-to-peer systems in a way that they could detect and respond to unexpected errors while still passing and recording real messages correctly. This actually remained an unsolved head-scratcher through to the late 1970s when the first major breakthrough occurred in distributed computing.

The problem with unexpected errors is just that: they’re unexpected. Some error types can be predicted and controlled. But if something unexpected happens, like a cosmic ray flipping the wrong bit in memory, one peer computer could send out two contradictory messages to the rest of the network. So how does a node detect this type of fault and manage it?

This type of fault became known as a byzantine fault as a result of NASA-funded work undertaken in 1978 and formally published in 1982 as The Byzantine Generals’ Problem by three leading computer scientists of the time. Thus began the race to develop ever better byzantine fault tolerant (BFT) consensus algorithms over the next decade, with obvious benefits for mission-critical systems in aircraft and military applications. In fact, every time you fly, you put your trust in BFT systems.

The next large breakthrough came in 1999 with the development of the practical byzantine fault tolerant (PBFT) consensus algorithm which was simpler and faster than the algorithms that had come before. So with all this history and momentum, why was the invention of blockchain so special?

Why blockchain?

Prior to blockchain, the world of consensus algorithms was stuck on the issue of identity. For a start, all the existing algorithms assumed you’d know the people you were planning on exchanging data with. Or at least their computers. So there was already a level of semi-trust in the set-up of your system. And, secondly, the algorithms required a complex series of message-passing between all participants before you could finally agree on any data. So the messaging complexity didn’t scale in a 1:1 manner with adding new nodes, instead it grew exponentially.

Blockchain came and blew all of that out of the water. No more trusted set-up or foreknowledge of the other people on the network, no problems with adding thousands or millions of other participants. Now you could share a common record of transactions, of history, of the truth, between millions of participants, without having to trust anyone. Just the maths behind the software. And if anyone tried to lie or alter the evidence, the whole network would say ‘no’.

The Cambrian explosion

Fast forward the next few years from Satoshi’s release of the white paper in 2008, through to the first release of the bitcoin software in 2009, the first pizza bought for bitcoins in 2010, price parity with an ounce of gold in 2013, the release of ethereum in 2015, and then an increasingly overwhelming array of start-ups and scams with the initial coin offerings (ICOs) of 2017.

Blockchain has gone up and over Gartner’s hype cycle faster than any other technology before it. From the revolutionary promises of global grassroots currencies grown from the ground up, free trade between rich and poor nations, restoring power to the people by removing all middlemen, blockchain has seen it all. It’s been truly amazing to watch the variety of ideas people want to build once they have a shared, irrevocable source of truth.

Any descriptions of the state of blockchain today will already be out of date by the time you read this book, so I won’t try. Just to say that the sector seems to be divided into four main categories of companies:

• Ledger builders: companies or groups building new distributed ledgers.
• dApp builders: distributed applications in which both the software and the data are stored and distributed according to a BFT consensus process between peers.
• Service providers: such as consultancies, marketing and specialised banking, providing services to companies in one of the first two categories.
• Centralized systems: custodial services such as cryptocurrency exchanges where you interact with the blockchain through a trusted party

Within the first category, there are a vast number of blockchains currently sitting comfortably alongside a smaller number of BFT systems built on the older consensus algorithms (R3’s Corda and Hyperledger Sawtooth to name a couple). People divide these various BFT distributed ledger systems into categories such as public and private, trusted and trustless, but the distinctions aren’t relevant here.

Many companies of the first two groups are in a period of rapid experimentation trying to tackle complex multi-user and multi-party problems, and prove that distributing the system makes it more resilient against attacks and failures, improving the relationship between individuals and companies.

The topics being explored across the industry are incredibly broad and well matched to the times in which we live: improving sustainability of supply chains, better sharing of healthcare data, more responsive corporate management, better provision of local government services and even safer air traffic control.

The road from here

Ask yourself this: had you heard of byzantine faults, distributed systems or consensus before you encountered blockchain? For the vast majority of people, the answer will be no. These subjects were out there, being researched by small teams dotted around the world, but generally consigned to footnotes in computer science and certainly not making splashy headlines.

The near volcanic eruption of blockchain onto the world scene has taken the radical ideas of distribution, peer to peer and trust into places as diverse as banking, forestry and even space. Many of the early radicals in the blockchain world (early being an amusing term here because it was less than ten years ago) fell in love with the technology as a potential solution to many of the world’s perceived ills: the excesses of capitalism, the increasing divide between the haves and have-nots, the forthcoming environmental catastrophe. These same people are still around, innovating and continuing to strive for transformation.

Blockchain can’t be uninvented. The trust machine can’t be stopped. But the ideas of iconoclasm and rebuilding the world without institutions or middlemen have been tempered somewhat by the experiences of the real world.

Instead I see the coming world of blockchain divided into two phases. The first will have more mature experimentation with ever larger groups of companies and even governments coming together to explore what canonical, irrefutable shared data means to their everyday operations. This will be followed by a vast dying off of unsustainable or unsuitable use cases for the technology, followed by a proliferation of trust machines behind the scenes in almost every aspect of our lives. One can only hope that these hold up to the original ideals of ensuring a better life for everyone.

Dr Robert Learney is head of distributed systems at Digital Catapult. He is involved in developing new programmes to help groups of companies from multiple sectors explore the potential of this technology to unlock economic growth for the United Kingdom. Prior to joining the Catapult, Rob co-founded the Imperial College Centre for Cryptocurrency Research and Engineering in 2014, aiming to create a cross-disciplinary academic focal point for blockchain research in London, and has been following developments across this sector ever since.

2.
A REVOLUTION IN TRANSACTIONS

Thomas Hartmann and Elke Kunde at IBM highlight how four technical features of blockchain are combining to redefine how networks of partners and suppliers can now work together

Electronic distributed ledgers are still a fairly recent technology. They are revolutionizing the processing of transactions, especially in the form of blockchain, the most prominent member of this technology family, which is a shared, decentral and immutable register of transactions.

Shared

‘Shared’ in this context means that several parties agree to joint bookkeeping for assets of any kind, covering all and any changes in their state during a transaction. No one records their own representation of the transaction, exchanging point-to-point messages as it progresses.

Bilateral messaging works fairly well, as long as all parties ar...