The communication and computing infrastructure we engineers have built in the past several decades has revolutionised the world economy. One of the most significant consequences of this revolution is the emergence of the digital platform economy. Indeed, while the top ten companies in the world ten years ago were dominated by companies exploiting natural resources, they are now dominated by companies that are in the business of providing a digital platform to connect producers and consumers.
The vast scale of these platforms is enabled by the underlying communication infrastructure connecting large numbers of individuals distributed around the world. However, the single company that runs a platform acts as a centralised intermediary. This is rather paradoxical as an original intent of the Internet was to allow a larger degree of decentralisation in our society.
The centralisation of platforms provides two significant benefits. First is scalability; it is easier to engineer and optimise a centralised system. Second is trust; society is already conditioned to trust large centralised entities. However, events in recent years have seen that there are significant limits to such centralised trust, imposed by frictions such as data centralisation, security centralisation, power centralisation, and competence centralisation. A challenge is whether one can build a system that is scalable and at the same maintain trust without the need of a centralised intermediary.
The communication industry is built around serving a basic human need. Just like communication, trust is also a basic human need, essential for a society to function and thrive. In fact, one of the main goals of communication is to acquire trust, so arguably trust is an even more basic need than communication.
Historically, trust has been provided by large centralised entities. Two breakthroughs that occurred in the past decade suggest the possibility of decentralised trust. First is the invention of Bitcoin by Satoshi Nakamoto in 2008. Most people know Bitcoin as the first cryptocurrency, but at the heart of it is the blockchain technology that provides trust by maintaining a decentralised immutable ledger of all past transactions.
Second is the invention of Ethereum in 2014, which shows that the blockchain technology invented by Nakamoto can be used not only to run a decentralised ledger for a currency, but also any Turing-complete program, so-called smart contracts. This vastly enlarges the range of applications of blockchains, far beyond just running cryptocurrencies. A broad vision is to build a single decentralised blockchain platform on which all digital platforms can be run—a platform for all platforms.
“To enable decentralised systems to compete with centralised systems, a core challenge is to design blockchain protocols that are scalable with the communication, computing, and storage resources of the underlying infrastructure, but at the same time retain the decentralisation and security benefits of Bitcoin and Ethereum”
We are still far away from this vision. Bitcoin and Ethereum provide decentralised trust, but the trust is not scalable. More precisely, it is not scalable with the communication and computing infrastructure on which these systems are run. For example, Bitcoin’s throughput is about seven transactions per second, translating to about 20 kbits per second, far below the communication bandwidths of modern systems. Moreover, in Bitcoin and Ethereum, the entire blockchain is maintained and stored at every node of the network. Hence, their performance does not scale with increasing computing and storage resources as more nodes participate in the network.
To enable decentralised systems to compete with centralised systems, a core challenge is to design blockchain protocols that are scalable with the communication, computing, and storage resources of the underlying infrastructure, but at the same time retain the decentralisation and security benefits of Bitcoin and Ethereum. In the past few years, many blockchain projects have emerged to take on this challenge, but no single scalable blockchain platform has yet emerged victorious. In fact, a folklore theorem, called the blockchain trilemma, states that it is impossible to build a system that simultaneously achieves decentralisation, security, and scalability.
Having been active contributors to the technology behind the wireless revolution, the state of affairs of blockchain technology development reminds the authors of the days of the development of the 3G cellular standard. In those good old days, multiple proposals emerged to compete to be the defining cellular standard. The competition was severe, but the adrenaline of the competition led to rapid progress in technology development.
We are seeing the same phenomenon here in blockchain infrastructure development. If Bitcoin is the first generation of blockchains and Ethereum is the second generation, then the game is on to compete for the third generation. A few of us from University of Illinois at Urbana-Champaign, Stanford, MIT, the University of Washington, and Carnegie-Mellon University have formed a team to participate in this competition. After working on this problem intensively, we believe we have solved the blockchain trilemma.
The beauty of Nakamoto’s blockchain consensus protocol—which powers Bitcoin, Ethereum, and many other blockchains—is its simplicity. All of the proposals that attempt to solve the blockchain trilemma depart from Nakamoto’s approach and involve very complex consensus protocols. These protocols are hard to implement, harder to incentivise, and their claimed security guarantees are either much worse than that of Bitcoin and Ethereum or have not been proven at all. Indeed, there are quite a few protocols that claimed to have the same security as Bitcoin, only to have vulnerabilities found later.
What we have discovered is a very natural generalisation of Nakamoto’s simple protocol that retains its simplicity and its strong security guarantees, but is at the same time scalable in the available communication, computing, and storage resources. Initial experimental results are very encouraging, but there is still a lot of work to do to build decentralised platforms around the ideas. Our long-term vision is that the protocol will be so resource-efficient that it can be run on phones, so that billions of users can participate in running the blockchain: the ultimate in decentralisation of trust.
Decades of communication and computing infrastructure building have led to the emergence of the digital platform economy. The next frontier is to build a scalable, decentralised trust infrastructure on top of this physical infrastructure so that the decentralised nature of the Internet can finally be realised.
About the author: David Tse is a member of the U.S. National Academy of Engineering, recipient of the IEEE Information Theory Society’s Claude E. Shannon Award, and the Thomas Kailath and Guanghan Xu Professor of Engineering at Stanford University. He also received the IEEE Richard W. Hamming Medal at the2019 IEEE Vision, Innovation, and Challenges Summit. He is the inventor of the proportional-fair scheduling algorithm used in all third and fourth-generation cellular systems, serving 2.7 billion subscribers around the world. The article features contributions from Pramod Viswanath.
The authors thank Sreeram Kannan and Vivek Bagaria for many discussions.
Interested in hearing more in person? Find out more at the Blockchain Expo World Series, Global, Europe and North America.