Investing in blockchain alternatives Pt I: Hedera Hashgraph

IOTA’s ‘Tangle’ is the most well-known project built on a direct acyclical graph (DAG), the most common blockchain alternative with the aim of creating a more scalable, energy efficient and time efficient network than the classical blockchain.

The protocol de jour appears to be EOS/Dan Larimer’s delegated proof-of-stake (DPoS) which has spawned other projects and also hybrid protocols like resulted delegated-proof-of-stake. Tron’s proof-of-authority and Neo’s delegated byzantine fault tolerance protocol also work on a similar premise to DPoS. In fact, it almost seems the more esoteric-sounding acronyms a project can fit into its whitepaper the more credibility it garners.

Hashgraph “the blockchain killer”

Hashgraph is distributed ledger technology built on its own version of a DAG, similar to IOTA, for the age of the Internet of Things. The technologies behind this self-styled “blockchain killer” are reiterations of two thirty-year-old simple computer science algorithms, the gossip and voting protocol, which Hashgraph has reinvented as “gossip about gossip” and “virtual voting”.

The technology is secured by the so-called gold standard of security, asynchronous Byzantine fault tolerance hashing, its algorithm requires no proof-of-work mining and it’s claimed it can handle hundreds of thousands of transactions per second, dwarfing even Visa.

According to the whitepaper, where Hashgraph and bitcoin differ in protocol is the efficiency in mining blocks. It posits that because a member of bitcoin’s network never knows for sure when consensus has been achieved – only a probability of confidence that rises over time – mining must be slowed down through PoW, so that if a chain does branch (fork) multiple times there is sufficiently long enough time between blocks to agree which is the real chain and to prune off the other branches.

“The hashgraph consensus algorithm is equivalent to a blockchain in which the ‘chain’ is constantly branching, without any pruning, where no blocks are ever stale, and where each miner is allowed to mine many new blocks per second, without proof-of-work, and with 100% efficiency.”       – Hashgraph Whitepaper

Gossip about Gossip and Virtual Voting

The inventor of the Hashgraph algorithm and co-founder of its parent company Swirlds is Leemon Baird, a former professor of IT and multi-tech entrepreneur who received his computer science Phd from Carnegie Mellon University in under three years – the fastest on record.

Baird describes the gossip about gossip protocol as a way to virally spread a great deal of information within the hashgraphs (packets of information): “If a new transaction is placed in the payload of an event, it will quickly spread to all members, until every member knows it. Alice will learn of the transaction. And she will know exactly when Bob learned of the transaction. And she will know exactly when Carol learned of the fact that Bob had learned of that transaction… and so on.”

Hashgraph’s virtual voting is proxy voting taken to the next level, where essentially no votes are cast at all over the network because stakeholder votes are inferred from their conversational history contained in the hashgraphs. The idea is that through this predictive virtual voting and gossip about gossip there is little algorithmic computing power being used and zero bytes of internet usage.

How to invest in hashgraph technology

As there has been no native hashgraph cryptocurrency, investing in this technology so far has been difficult for those who aren’t accredited investors and the only vehicle to do so was Swirlds, a private company. This March, Baird and Swirlds co-founder Mance Harmon unveiled Hedera Hashgraph, a new company dedicated to developing a distributed public network built upon hashgraph.

Hedera will have its own utility token, but the pre-sale will only be open to accredited US investors and there will be no public ICO. Participants of the network will eventually be able earn the token once the network is up and running which can be used to access applications on the network or staked to run as a node to provide security to the network. However, projects built on hashgraph can issue their own ICO and gaming platforms. Satori’s upcoming LIT token issuance will potentially be the first to ICO on hashgraph.

Another interesting application for hashgraph is for smart contracts settlement. Last year alone half a billion dollars was lost in smart contracts transactions, the most notorious being the bug in the code of the Parity wallet that resulted in $300m of ether being destroyed.

Sagewise is building a platform to integrate human adjudication with smart contracts on the hashgraph network, giving parties involved in the self-executing contracts the ability to monitor, freeze and resolve disputes.

“One of the bigger issues with smart contracts is that even if there is not an explicit bug, there might just be something in the contract that gives a single party a massive amount of control. We’ve seen several ICO contracts that allow the creator to freeze or in some cases even execute custom code inside the smart contract. This can defeat the purpose of using a decentralized smart contract completely and is the reason we provide tools that allow for more responsible governance of individual contracts,” said Daniel Rice, co-founder of Sagewise.

Rice said the company is considering an ICO, but nothing official has been planned.

“The fourth generation blockchain”

Hashgraph is challenging for the coveted spot of being the world’s first ‘mass-adopted’ public distributed ledger, with utility to encompass currencies, financial services, healthcare, IoT, gaming and nearly every other consumer touchpoint.

Accordingly, the Hashgraph team refers to itself as the “fourth generation of blockchain” in the epochs of DLT:

1. Bitcoin and altcoins: the transfer of digital currency
2. Blockchain: storing ownership of items such as land, music, stocks on a blockchain
3. Ethereum: the use of smart contracts as agreements between buyers and sellers
4. Hashgraph: fair matching of buyers and sellers to execute smart contracts and Dapps

Distributed or decentralized? Public or permissioned?

Unlike most in the DLT space the Hashgraph algorithm is not open-source but patented by Swirlds and its governance is overseen by the Hedera Hashgraph Council – a distributed group “of up to 39 renowned enterprises and organizations across multiple industries and geographies”.

Hashgraph will be governed a board of corporations in a system akin to that of Visa. The Hedera foundation will hold the majority of voting supply (tokens), which isn’t ideal for stakeholder voting power but Hedera claims its control of the token supply is to prevent entities from gaining too much control of the network while its cheap. However it’s not even clear how the corporate entities are chosen or how the vetting procedure works, but all will serve two three-year terms – apart from hashgraph parent company Swirlds which will have an indefinite tenure on the council.

Interestingly, in the entirety of the Hashgraph whitepaper nowhere do the words public, private, permissioned or decentralized appear, but the emphasis is on its “distributed” nature, which sounds like a euphemism for permissioned. So it seems strange to read on the Hedera website “The hashgraph algorithm is highly decentralized – there are no leaders, miners, coordinators or block producers with special influence towards consensus. Separately, the Hedera governance model is also decentralized.”

Also Herdera’s own description of the role of the governing council seems to jibe with decentralization in the sense of top-down rule setting: “The elected Governing Board will govern the council by establishing policy for council membership, regulating the network rules and tokens, and approving changes to the platform codebase.”

It sounds like Hedera is trying to do something akin to R3 with its Corda blockchain, and there is plenty of discussion around its merits.

Conclusion

While Hashgraph’s DAG infrastructure and speed sounds academically impressive on paper, its performance may suffer due to its high latency of 3 seconds – for example, its blockchain counterpart, EOS, has 0.5s latency. Whether this latency difference matters in an asynchronous network is yet to be seen in a real world simulation.

EOS’ Dan Larimer said: “There is a difference in quantity and quality, validated versus unvalidated. If all you’re trying to do is order events without consideration of validity Hashgraph can work, but if you need to do dependent relationships between them I don’t think it’s so scalable.”

It also sounds like it has swapped out EOS’ “delegated block producers” for “distributed governance” and replaced them with the interests of corporations, which is anathema to the decentralized ethos but will be attractive to legacy world hegemonies that want to make the transition to the top of the “distributed” world.

As an investment opportunity, retail or not, this is something to pay close attention to.

Follow @AndrewBNC

Within the myriad permutations of blockchain protocols the battle is no longer between the duopoly of proof-of-work and proof-of-stake but a universe of spin-offs, hybrids and alternatives. Now even blockchain technology itself faces disruption by distributed ledger technologies that forgo the blockchain stack and architecture. In this first part of a series we’ll look at alternative blockchain projects and how to invest in them.

Subscribe to BNC’s newsletters for insights and forecasts direct to your inbox