Evaluating the State of Distributed Ledger Technology

Two leading experts share their takes on what’s realistically next for digital currencies

By Alicia Roisman-Ismach

Digital currencies are making waves, but will Bitcoin (BTC) and its crypto-brethren ever replace the payments system as we know it? For Ari Juels and Ittay Eyal, co-director and associate director, respectively, of the Initiative for CryptoCurrencies and Contracts (IC3), cryptocurrencies and the distributed ledger technology that underpins them face several substantial limitations before they can reach mainstream adoption. But the IC3 is working hard to address them and usher in the next generation of distributed ledger technology. Transaction Trends asked them to share some of their observations.

What needs to happen to reduce the volatility of BTC and other currencies?

Ari Juels: I don’t know about BTC. It has several forms of built-in instability. Its abhorrent and unsustainable guzzling of electricity is one. Economist Alex de Vries estimates that Bitcoin mining will consume 7.7 gigawatts (GW) by the end of 2018. Fractious governance is another. Poor scaling is yet another, and I’m not optimistic about payment channels and other scaling solutions. So, it’s hard for me to see volatility subside much.

Ittay Eyal: There are some obvious necessary (though probably not sufficient) elements:

• The market size becomes sufficiently large to make manipulation too expensive.
• The currency continues working consistently, without bugs, and through difficult governance events. This would demonstrate the stability of the infrastructure.
• Adoption reaches a stable point. Once it takes a stable share of payment markets, or even just investments, exchange rate growth could calm.

When do you think the general population will own cryptocurrencies?

Eyal: Once they are reliably available from trusted sources, which could be banks or even convenience stores selling cryptocurrency gift cards. But this is just ownership. Once ownership becomes easy and the exchange rate stabilizes, then it will be much easier to adopt for actual payments, and beyond.

Juels: I prefer to keep hoping it won’t. Well-vetted, institutionally custodied tokens and cryptocurrencies might find a place in investment portfolios, and it wouldn’t surprise me to see that happen in the next five years. But then the cryptocurrency dream of true decentralization will essentially be toothless.

What key technical milestones would have to occur in order for digital currency to replace traditional currency/payments?

Juels: In a sense, it already has [replaced traditional currency]. Most currency has been dematerialized. If you mean blockchain-based currency, I think that the active embrace of governments will be necessary, and that will be a quagmire. While conceived by libertarians, cryptocurrency can alternatively be a powerful tool of repression in the hands of governments. Without careful and intentional safeguards for user privacy, it can enable pervasive and automatic surveillance of monetary transactions.

Eyal: Current digital currency systems face several technical limitations that prevent them from being able to handle some significant portion of the world’s payments. The following questions are under intensive study by the research and development communities.

• Scalability: Our work (dubbed Bitcoin-NG, or Next Generation, though it’s a general construction not just for Bitcoin) showed that the blockchain protocol need not limit throughput. However, the question of how to shard the blockchain (without reducing security) such that it can scale beyond the capacity of a single machine remains open.
• Proof-of-Work (PoW): The environmental impact of PoW mining is problematic. The most advanced alternative, proof-of-stake (PoS), works under significantly stronger assumptions. (PoS assumes the majority of nodes in a network are altruistic, whereas PoW assumes all nodes are rational.)
• Client security: The mechanisms for securing client keys are insufficient. This applies both to individuals storing their personal funds and to companies storing more significant amounts.

Juels: The IC3 Grand Challenges, for the most part, reflect a consensus among our faculty on the major technical obstacles the community will need to overcome for blockchains to see wide and fruitful deployment.

Scaling is attracting the most attention. Additionally, it’s relatively easy to achieve if you’re willing to weaken the strong trust model embodied in fully permissionless systems, and move toward more centralization, which I think is appropriate in many settings.

Our sixth and newest challenge—“Sound Migration”—represents a problem that one of our industry colleagues (Amber Baldet, then of JPMorgan Chase) sensitized us to. It may well be the most important, because without accommodating legacy systems, blockchains are going to be very slow to emerge. This is something we failed to appreciate but definitely do now. It’s insights like these that reflect the value IC3’s industry partners bring to our research agenda, and IC3 more generally. We can’t simply jettison existing systems and start from scratch. Legacy-system integration is hard but necessary.

Is there a tradeoff between transparency and confidentiality in cryptocurrencies and distributed ledgers? How do you strike a balance? And how do you protect user privacy?

Eyal: Indeed, there is a tough tradeoff, with different entities expecting very different properties. Between law enforcement bodies that would probably opt for complete transparency and individuals that would keep their privacy, there are entities such as banks and insurance companies that might want something in between. In practice, I believe a lot of the cryptocurrencies will still be stored in banks, so the situation would not be too far from what we have today—for better or worse.

Juels: There’s a strong tension between the two. Striking a balance isn’t easy. There are some settings in which cryptography, and zero-knowledge proofs and/or secure multiparty computation, are helpful, but they are only practical in niche settings. This is why I’m more bullish on trusted hardware, for all its failings (which are not inconsiderable).

Aside from payments, where might there be some applications for blockchain within the financial space?

Eyal: The programmatic nature of blockchains allows for going beyond payments to achieve more elaborate constructs. This includes limited-trust escrow and custody services, atomic swap mechanisms and other financial tools, but also games and prediction markets. Using oracles—which are systems for relaying off-chain, real-world data onto the blockchain in a trustworthy way—can allow for additional services relating to external data such as financial information or insurance events. But serious usage still seems a bit further down the road due to the volatile exchange rate of cryptocurrencies.

Juels: Smart contracts enable the creation of a whole range of financial instruments, from automated versions of existing instruments to wholly new ones. Token mania has already demonstrated the sea change that this technology can precipitate.

“Smart contracts” are only as good as the information they have. How can we set up data pipelines to verify the information that goes into smart contracts?

Juels: We’ve had an oracle called Town Crier [www.town-crier.org] running in Ethereum for over a year. The tamperproof nature of the hardware allows data to be transmitted faithfully from websites to blockchains and can also provide confidentiality for queries. Oracles themselves can also be decentralized along several dimensions, which can enhance trust.

Eyal: There are already solutions utilizing trusted hardware (like Town Crier) and game-theoretical constructions that motivate participants to enter accurate data into the blockchain (like Schelling).

What do you want businesses, and specifically payments companies, to know about distributed ledgers (DLs)? What should they consider before adopting them?

Juels: Blockchain technology may prove most valuable as a catalyst for the rediscovery of extremely powerful, but underappreciated, technologies and disciplines, like applied cryptography, fault-tolerant systems, high-assurance software engineering, and more.

DLs are not magic pixie dust. They solve some problems, but they are being hyped as solutions to challenges they cannot resolve on their own. It’s important for those contemplating adoption to obtain objective insight into the technology landscape and the promises of cutting-edge research.