A $3.3 Billion Claim: Has Cardano's Blockchain 'Solved' Proof-of-Stake?
Charles Hoskinson explains cardano's iterative approach to security, and how its interaction with academic gives it an advantage over other designs.
"Proof-of-stake is solved."
Even in an industry that's seen no shortage of grand proclamations, those words, issued by entrepreneur Charles Hoskinson in April, grabbed attention.
The CEO of blockchain firm IOHK (and one-time CEO of Ethereum), Hoskinson was seeking to send a message about a new research paper, one he believes proved that the company's novel twist on how blockchains come to consensus, called Ouroboros, had addressed long-standing concerns about whether the model can sufficiently secure investor funds.
Given the size of the claim (and its impact), it's an assertion that sparked doubt from cardano's more prominent competitors. However, months later the team at IOHK maintain Ouroboros may be the answer to one of crypto's most divisive questions – whether so-called proof-of-stake systems offer solutions to some of the industry's pressing problems.
So far, the market appears to be interested in the opportunity to support the thesis.
Soon to power the public blockchain Cardano, Ouroboros may one day support the world's eighth-largest cryptocurrency, with its 25 billion ADA tokens worth $3.3 billion. And a look at the history of proof-of-stake systems perhaps showcases why so much money is on the line.
First pitched by developers Scott Nadal and Sunny King in 2012, proof-of-stake offers what is claimed to be a more sustainable alternative to proof-of-work, the consensus method underlying the world's biggest blockchain by market cap, Bitcoin.
Allowing users to vote or "stake" their coins on a transaction history in exchange for rewards (instead of burning computational energy), it's relatively untested, having so far only been adopted in hybrid, small-scale or delegated formats.
So, while Bitcoin's security is comparatively proven (its blockchain is currently sustaining $114 billion and has held up for years), many crypto coders believe proof-of-stake is necessary to transition the industry into the next phase, in which users no longer have to own hardware in order to claim a blockchain's rewards.
The trouble is, no one can agree on how this should be done.
"Different consensus algorithms do well in different environments," Nate Rush, a proof-of-stake researcher for Ethereum, told CoinDesk, "If the assumptions that some protocol is 'solved' under turn out to break or be unrealistic, then this protocol can fail."
Still, the team behind Cardano, IOHK, have worked to secure academic partnerships, as well as relationships with researchers in the field distributed computation in an effort to prove the proof-of-stake model can be achieved.
Taking the security of Bitcoin as its starting point, the chief scientist behind the protocol, Aggelos Kiayias, has created formal proofs for each step of the protocol's design, used to dispel doubt as to the algorithm's ability to protect assets.
Kiayias told CoinDesk:
Peer-reviewed protocol
But it's not just proof-of-stake researchers that disagree – in the broader landscape of consensus design, there are some who believe proof-of-stake is doomed from the start.
For example, Dahlia Malkhi, a distributed system researcher, claimed earlier this year that Ethereum's proof-of-stake model, Casper, is "fundamentally vulnerable"– leading to a system where consensus is powered by the wealthy.
Yet, in this atmosphere of skepticism, Cardano has amassed a high degree of academic support, building strategic university relationships through IOHK – a for-profit company with centers in several universities, including the Toyko Institute of technology in Japan, University of Athens, and University of Edinburgh.
"We basically have a kind of a university-company relationship, so what we generally do is we set up research centers, we embed some IOHK personnel within those research centers, and we subsidize the lab, and then we have some sort of control or influence over the research agenda," Hoskinson said.
Responsible for Ethereum Classic and Zencash, as well as Cardano itself, according to Hoskinson, these academic partnerships feed back into the IOHK's cryptocurrency solutions.
"Generally, the output of this relationship are peer-reviewed papers and we have a team within IOHK that take those papers, extract specifications from them and then put them in the pipeline for implementation into products," Hoskinson continued.
Alongside further relationships at Lancaster, Kent University, Oxford, and Illinois, the advantage is that the relatively small pool of researchers equipped in the topic have had their eyes on Ouroborus, which has been toured around at various academic conferences as well.
"We've brought in the total set of people who are actually looking at trying to break, trying to enhance, our core protocols, and there's a lot of hard work that has been achieved as a result of that," Hoskinson told CoinDesk.
Gearing up to become a fully fledged smart contract and cryptocurrency platform, Cardano currently has limited functionality, but it is introducing the necessary features to transition to proof-of-stake throughout this year.
Currently tweaking the final details under simulations, Hoskinson urged that Ouroborus comes with advantages over other protocols. For example, he says the system is the only one that will allow users to stake from cold storage, and use multiple addresses to manage their finances.
Hoskinson told CoinDesk:
Theoretical unknowns
Still, at the time of writing, it's unclear how the protocol will behave in the wild, and there are ways in which the wider proof-of-stake research community hasn't been entirely receptive of Cardano's claims.
For example, no one has been as critical as EOS's creator Dan Larimer, a former colleague of Hoskinson, who wrote that not only is "Ouroboros is a 400-pound bulletproof vest that doesn't actually stop the real bullets," but claimed it was a badly conceived variant of an algorithm he had designed in 2014.
(EOS uses a form of proof-of-stake that relies on delegated nodes that have been nominated the task of reaching consensus.)
Ethereum, of which Hoskinson was a co-founder, has also signalled skepticism.
For example, speaking on Reddit, Ethereum founder Vitalik Buterin warned one security assumption of the blockchain could allow hacker to create false histories on the blockchain.
There's also different approaches to design. While the protocol itself is in place, IOHK are still working on building the underlying incentive scheme, something that Casper researcher Vlad Zamfir believes should be designed in tandem with the tech.
Attempts by CoinDesk to engage researchers with published work on proof-of-stake also returned mixed results, with several offering no comment or suggesting they hadn't yet looked into the technology and its specific claims.
Theory meets reality
But according to Emin Gün Sirer, a Cornell University professor and researcher in consensus protocols, this is typical of the field.
"Ouroboros has the advantage that it is peer reviewed and a well-credentialed research group stands behind the effort," Gün Sirer told CoinDesk, "But it also suffers from a downside that plague many early proof-of-stake protocols, namely: the papers are long, dense and full of subtle proofs."
As a result, Gün Sirer said, "No two researchers in this area seem to agree on which papers have valid proofs and which have redefined the properties so as to make proofs meaningless. The academic community vets papers for academic rigor, not real world application."
That said, Hoskinson believes this interaction with academia is essential for cryptocurrencies to migrate onto the next phase. He warned that while many blockchains make wide claims for the scalability and security of their product, there's not many laymen with the skillset to properly evaluate those claims.
As such, Hoskinson said he anticipates more research on Ouroboros and Cardano to emerge going forward.
"We all talk about the Tendermints and the Decreds and the Caspers because these are industry accepted ones and they have good marketing," Hoskinson said.
However, he continued:
STORY CONTINUES BELOW
Rubix cube via Shutterstock