A veteran researcher behind two influential papers in the emerging field of crypto-economics is gearing up to launch a new cryptocurrency.
Revealed in an exclusive interview with CoinDesk, Yonatan Sompolinsky, co-author of the GHOST and SPECTRE protocol alongside Dr. Aviv Zohar, intends to release the as-yet unnamed project in late 2018. Based on his body of work, the project aims to solve one of the industry's biggest challenges: the speed at which transactions are disseminated and recorded by distributed ledger systems.
Founders of the project include SPECTRE co-author Yoad Lewenberg, and researcher Ethan Hileman, who worked with the team behind bitcoin privacy project TumbleBit. However, the team is not only equipped with developers, but also business experts - including Guy Corem, the former CEO of Israeli-based bitcoin mining firm Spondoolies-Tech, which raised more than $12 million in venture funding before shuttering operations.
Together, the team has founded a new startup called DAGlabs, which is said to be raising $15 million as part of a Series A round.
Yet, it's the vision for the project that is perhaps most notable, given how it intends to achieve its goal.
Rather than using a blockchain system, the public cryptocurrency will be among the first to operate via a system called a direct acyclic graph (DAG) - technology that Sompolinsky framed as a way to finally create a viable payments rail with distributed ledger concepts.
Sompolinsky told CoinDesk:
"We want to change blockchain into blockDAG. Whether you buy coins or not, you should think about this technology as the next step, releasing the blockchain from the naivety of the chain structure. You will feel this is the natural step towards using a real system."
For all the innovation, however, also notable is what the new cryptocurrency will share with bitcoin, today the largest and most valuable blockchain system, but one that is seen as increasingly less relevant for payments.
According to the DAGlabs team, the new cryptocurrency will use a proof-of-work mining algorithm in which anyone who purchases hardware can compete for its rewards. That said, Sompolinsky doesn't see bitcoin and other proof-of-work-based cryptocurrencies as competitors, so to speak.
"I'm not competing with this entire plethora of new blockchains. We have a very boutique and niche and specific one and a well-defined one. We want to scale up layer one," he said.
Record of science
As told by Sompolinsky, the project is also a critique on the current state of blockchain development, which he views as having been held back by in-fighting.
In particular, he cited his experience with arguments about capacity levels on bitcoin, in which developers have often tried to push solutions to scalability in a way that did not change or update the first layer of the system - the blockchain itself. Referring to last year's Satoshi Roundtable summit, an invite-only event in Cancun, Mexico, he recalled an experience where he was shocked by the state of the conversation.
"Cancun was an eye-opener for me. In Cancun, everyone was fighting about 1 MB to 2 MB ... no one was talking about increasing on-chain scalability by orders of magnitude," he said.
Sompolinsky framed the new cryptocurrency as a "vehicle" to enable researchers to take the next step in evaluating this line of exploration, one he argued will benefit from being tested under open market conditions.
But given the slate of open-source projects utilizing the initial coin offering (ICO) model as a way to solicit market funding, often for untested concepts, he was also keen to differentiate his project as one based on years of accumulated research.
"There are not 800 projects that implement a concept like DAG. Very, very few, maybe less than five that I know of, try to scale up the layer one. There are not 800 projects that say we should abandon the concept of the chain in favor of a graph of blocks," he said.
Among those that do are IOTA and byteball, the former being one of the few cryptocurrencies yet to garner a total market capitalization of more than $1 billion.
When complete, the final network should appeal to anyone who wants to use a cryptocurrency with "very fast confirmation and low fees," although that's not to say the concept has been perfected. According to Sompolinsky, there are still plenty of items up for debate.
For example, he is still open-minded about how he will structure any issuance, stating he is most strongly considering the model used by Zcash Company - the creators of the zcash protocol - wherein accredited investors are given tokens in stages for their support.
How it works
Still, even those experienced in the field of cryptocurrency may find the concept odd. After all, after the hype around cryptocurrency faded, the so-called "underlying blockchain technology" was often touted as the real secret sauce.
That notable researchers would stake a counter-thesis then, is of interest, though it's arguably been a development that has simply received less attention over the years.
Explored by Sompolinsky since his earliest work, the idea is that the process of ordering transactions into blocks, then selecting one to add to the chain, could be better optimized. In the SPECTRE concept, blocks are created at the rate of about 10 per second (as opposed to, say, ethereum where one block is created every 21 seconds). All of these blocks are referenced in a DAG, and multiple, interwoven threads of blocks are created. Then, the most valid transaction history is "voted" in by miners selecting the most inter-referential block graph.
And because this allows for such a high quantity of transactions to be performed on the network, the transaction history won't be permanently stored. Instead, transaction history will only be stored for a limited amount of time, and once it has been validated, will be removed.
Other barriers, including backbone congestions and bandwidth, could lie ahead, though. The team intends to address these potential issues by building incentives into the protocol that encourage participating agents to behave correctly.
Still, if the concept on the whole sounds familiar, you've likely come across it when reading about ethereum, which incorporated some of these ideas in how it rewards so-called "uncle blocks" - those that are not selected for inclusion in the ethereum blockchain, but still include useful work.
Although uncles aren't considered ultimately valid, they're still profitable to mine, and still get referenced in the blockchain ledger itself.
Indeed, what might be most notable about the SPECTRE project is that Sompolinsky's ideas have often proved to be of influence. For example, Casper, ethereum's hotly debated proof-of-stake protocol, derives its name from the GHOST protocol (a play on the "Casper the Friendly Ghost" cartoon series).
While the bones of the project are in place, there's still a few questions that need to be fleshed out before the cryptocurrency can be taken to market.
Currently under development, the final protocol will be introduced in an upcoming white paper called SPECTRE2. The tech will then begin testing in autumn next year, and by winter, the new cryptocurrency will be launched. Ongoing updates will then continue apace.
For one, DAGlabs has been working on combining the SPECTRE cryptocurrency with MimbleWimble, the natively private cryptocurrency that has been the cause of much excitement in the community. Further down the line, the platform also wants to allow DAG-based smart contracts to be written into the platform.
But the final step that DAGlabs is pivoting towards is perhaps as ambitious as the cryptocurrency itself. Once the team has established a functioning cryptocurrency, they intend to allow the DAG infrastructure to facilitate something along the lines of "merged mining," which is when the underlying hardware can be used to support a number of cryptocurrencies simultaneously.
Still, it's the manner in which Sompolinsky will launch the project that he ultimately wants to call attention to - calling the years of research and engagement with the open-source community, a "respectable" path to market.
"The easy path for us was to do an ICO; way easier than what we're doing now. We're going for the more respectable path of an equity."
Correction: Some quotes have been updated for clarity.
Disclosure: CoinDesk is a subsidiary of Digital Currency Group, which has an ownership stake in Zcash Company, the for-profit entity that develops the zcash protocol.
Scary balloons image via Shutterstock