WSO2 sponsored this post.
Blockchain promises to redefine trust. On one hand, it lets us build decentralized systems where we do not need to trust the owners or operators of the system. On the other hand, blockchain lets previously untrusted parties establish trust quickly and efficiently. This enables developers to build novel applications that can work in untrusted environments. At the same time, it is not readily apparent where blockchain use cases both are feasible and can deliver clear value. Therefore, my colleagues and I recently took a systematic approach to analyze blockchain using the Emerging Technology Analysis Canvas (ETAC), which takes a broad view of emerging technology by probing impact, feasibility, risks and future timelines.
It’s hard to talk about all the proposed blockchain use cases collectively and come to reliable conclusions. This is because they have different requirements and goals. Therefore, we began our analysis by surveying the many blockchain use cases and categorizing them under 10 categories as depicted in the following diagram.
The main blockchain use-case categories then let us identify common traits. For example, ledgers, new internet services, autonomous ecosystems and lightweight financial systems let us build decentralized systems without trusting the operators of the system. Meanwhile, disintermediation (removing brokers in between), provenance (tracking the history) and smart contracts allow exchanges of information or payments requiring fewer security checks.
New Versus Old Systems of Trust
Most of the use cases noted here have already been implemented in some way. So, why care about blockchain? The answer is that blockchain can provide a new kind of trust.
Traditional, pre-blockchain systems generally function as they should. However, they implicitly assume two kinds of trust. In the first instance, we trust the “super users” of what are centralized implementations. Here, one person or a few individuals have deep access to the system and are deemed trustworthy. Alternatively, with an organization or government, we can reasonably assume processes are in place that should deter wrongdoing (not all world governments fall under this category, of course).
The second way we establish trust is through an out-of-bound means, such as signing a legal contract, obtaining a reference or by providing a credit card to gain access. This is why most systems or ecosystems require you to provide credentials, which you need to create through some other channel, before you can access them.
Unlike traditional trust systems, blockchain-based systems can operate without either of the assumptions being true. Operating without the first assumption is decentralization and doing so without the second is known as “dynamic trust establishment.” However, our ability to operate without these assumptions and achieve a new level of trust does not always mean that we should.
We need to consider the tradeoff between the cost of using blockchain and the potential returns (both monetary as well as abstract value such as more trust) This is not a technical decision, but a consideration of our values and how much risk we’re willing to bear.
Weighing the Costs of Blockchain
Our analysis identified several challenges (technical limitations that will likely be fixed in the future) and risks (inherent aspects of blockchain unlikely to change). Blockchain challenges are limited scalability and latency, limited privacy, storage constraints and unsustainable consensus (e.g. current consensus algorithms are slow and consumes significant computing power). Meanwhile, blockchain risks include irrevocability, regulator absence, misunderstood side effects, fluctuations in bitcoin prices and unclear regulatory responses.
In our analysis, we then evaluated each use case category in the context of blockchain’s challenges and risks and we arrived at three levels of feasibility.
First, blockchain technology is ready for applications in digital currency, including initial coin offerings (ICOs); provenance, e.g. supply chains and other B2B scenarios; and disintermediation. We expect to see use cases in the next three years.
Second, ledgers (of identity, ownership, status and authority), voting and healthcare, are only feasible for limited use cases where the technical limitations do not hinder them.
With the remaining use cases — lightweight financial systems, smart contracts, new internet apps and autonomous ecosystems — blockchain faces significant challenges, including performance, irrevocability, the need for regulation and lack of census mechanisms. These are hard problems and it could take at least five to 10 years to find answers.
In most cases, today’s centralized or semi-centralized solutions for establishing trust are faster, have more throughput and are cheaper than decentralized blockchain-based solutions when considered holistically. Adopting blockchain means that we are willing to pay for added delays and duplications.
Is Blockchain Worth It?
Neither the decentralization nor dynamic trust establishment enabled by blockchain is free. However, while true decentralization is expensive, once in place, it makes dynamic trust establishment easy to implement.
Decentralization can be useful for the following scenarios:
- Limiting government censorship and control.
- Avoiding a single organization from controlling critical systems or data.
- Preventing a single or few rogue employees from making significant damages.
- Applying rules of the systems to everyone evenly (limit cases where people who have access to key people from the system enjoy special treatment).
- Reducing damage because few accounts are compromised.
So far, we’ve managed to run our systems without embracing decentralization scenarios.
Moreover, the polarized arguments around blockchain’s value suggest there is no shared understanding of the value of decentralization. People express concern about the arbitrary power of governments as well as large organizations. But do they understand tradeoffs as additional resources required to attain higher trust? Similarly, people worry about privacy, but most of us share data daily in exchange for free access to large internet companies and social media platforms.
Clearly, decentralization needs to be part of a policy decision that is taken only after wide discussion. One one hand, in an increasingly software-controlled world — from banking to healthcare and autonomous cars to name just a few — the risks associated with centralized systems are increasing. On the other hand, trying to attain full decentralization can kill blockchain by setting a too-ambitious target because the resulting cost will be too high.
Fortunately, centralized versus decentralized does not have to be an all-or-nothing decision. Multiple levels of decentralization are possible. For example, private blockchains essentially are semi-decentralized because any action requires consensus among a few key players. Therefore, it is important to critically examine each blockchain use case under three choices: centralized, semi-decentralized, decentralized.
Significant financial investments have been made in blockchain, but if the quest for a fully decentralized solution takes too long; it will put the future of blockchain at risk. Presumably, we could support semi-decentralized solutions (private blockchains) much faster than full decentralization. This makes a good case for starting with a semi-decentralized approach to a blockchain use case to minimize risk and then striving for full decentralization in a second phase.
Blockchain provides mechanisms for establishing trust that reduce the risks associated with centralized systems and enable agility by automating the verification required to establish trust. However, compared to current decentralized or semi-decentralized blockchain solutions, centralized solutions are faster, have more throughput and cost less to implement. That said, as governments and market demands address the technical challenges blockchain faces, the associated costs and barriers to implementation will be reduced. And we expect some categories of blockchain use cases to go into production within the next three years.
Deciding which blockchain use case to invest in and when requires a critical analysis of the tradeoffs. A more detailed analysis of blockchain using the ETAC framework and a discussion of these trade-offs can be found in our paper, “A Use Case Centric Survey of Blockchain: Status Quo and Future Directions.”
Feature image via Pixabay.