Apr 10, 2017
Defiled Consensus: Waiting for a Digital Messiah

Some anthropologists believe consensus is a universal decision-making system inherent in primitive societies where people struggled for survival and inclined towards equal partnership. Modern-day people who live in the era of representative democracy think that consensus is mostly related to mind-numbing teambuilding sessions. However, the world of blockchain brought back meaning to the issues of consensus.

Some anthropologists believe consensus is a universal decision-making system inherent in primitive societies where people struggled for survival and inclined towards equal partnership. Modern-day people who live in the era of representative democracy think that consensus is mostly related to mind-numbing teambuilding sessions. However, the world of blockchain brought back meaning to the issues of consensus.

Consensus and Status Quo

Consensus is a process of collective decision-making by obtaining consent from all involved parties. There is a difference between consensus and simple majority vote. The latter divides all participants into two obvious groups, being winners and losers. Consensus, on the other hand, is antagonistic to any conflict of interest and implies that everybody’s a winner. Even there’s a group not quite happy with the outcome, they may still give their consent because they heed the interests of the entire group and are willing to put common goal on top of their own.

No wonder that in the modern world, where communities may be enormous and people tend to be individualistic, consensus democracy isn’t blooming. Most modern-day democracies are based on majority principles that assume minorities has to obey majorities. The principle has proved itself efficient in dealing with socially polarizing issues plentiful in any nation state. A big individualist society just makes it impossible to base all decision on consensus. Not sliding back to the majority theory would suffice.

Still, consensus-based decision-making is an important tool of democracy commonly used in international structures that emphasize equal status of all participants and uses veto (a notable example is the U.N. Security Council).

Those monitoring such entities know that seeking a consensus in groups lacking mutual trust have a peculiar feature, which is the fact that this kind of decision-making is mostly focused on maintaining status quo due to being incapable of making radical decisions on complex issues concerning the group’s members. That’s why the mechanism is mostly applicable in groups seeking to maintain status quo.

The machine incarnation of consensus has the same feature. In trustless systems, machine consensus is mostly focused on maintaining status quo.

Machine Consensus

Blockchain first came to be due to an anonymous visionary programmer known as Satoshi Nakamoto. Over time, the technology ceased being a mere foundation of bitcoin’s payments system and went beyond its original purpose. However, its original mission was about maintaining user consensus and transaction immutability in a pseudonymous trustless network.

Bitcoin network and similar p2p solutions are the modern-day iteration of the Byzantine generals problem. It’s a classic cryptographic problem where the network of several nodes that have to optimize and synchronize their operations without mutual trust is allegorically represented by Byzantine generals going to storm an enemy stronghold. Any seasoned general certainly suspects other generals of betrayal, and some of them are indeed likely to commit treason. There has to be an algorithm enabling honest generals to reach a consensus notwithstanding the enemy workings.

In other words, when network nodes don’t trust each other, we have to develop rules of a game that would allow users to trust the entire system without trusting each other.

In major networks, common consensus without intermediaries is almost impossible. That’s where blockchain comes into play: by delegating consensus seeking to a machine algorithm, it settles the Byzantine problem for a network with unlimited number of members. According to digital industry stakeholders, blockchain is the best basis for global autonomous networks like the emerging Internet of Things.

However, back to bitcoin, where social consensus becomes different from machine one?

Proof-of-Work

It’s not that easy to recognize the features of intuitive social consensus in bitcoin’s consensus algorithm.

This algorithm known as Proof-of-Work allows miners to build blockchain under commonly accepted rules. The process is about racing to solve a complex cryptographic puzzle. The first to solve it has the honor of introducing a new block to the chain if most nodes of the network confirm the block is correct. Then the process starts all over again.

This mechanism does two things: sets up rules for the network and keeps the system from desynchronizing and forking. Even though the algorithm is competitive, its concept is about cooperation and maintaining network uniformity. Miners who start using a different set of rules play against the system and therefore automatically violate the consensus. They are de-facto purged from the system and lose their chances to earn money.

Machine consensus could also be considered a chain making users jointly responsible and causing them to agree with everything occurring in the system under the threat of expulsion. The nodes that don’t recognize a block accepted by the majority (or recognize a block discharged by the majority) are automatically removed from the network consensus field and form their own parallel chain.

Hardfork, by definition, exists beyond the old consensus and results in the need to find a new one. This becomes a problem only when the network has enough miners willing to take over the initiative of consensus-reaching or to go away into their own fork (or, in other words, when Byzantine generals collude).

Behemoths in a Sandbox

“The proof-of-work also solves the problem of determining representation in majority decision making. If the majority were based on one-IP-address-one-vote, it could be subverted by anyone able to allocate many IPs. Proof-of-work is essentially one-CPU-one-vote. The majority decision is represented by the longest chain, which has the greatest proof-of-work effort invested in it”.

Satoshi Nakamoto, Bitcoin WP

It wasn’t Satoshi who invented PoW. The algorithm first appeared back in the nineties as a protection against spam attacks. Originally, it wasn’t a profitable target for attacks, however, after bitcoin, when simple hashing has become a source of profit, it has become a problem.

In their race for profits, miners created major pools while entrepreneurs invested a fortune in ASIC farms. The consensus algorithm mutated into a gluttonous monster eating up enormous resources. What Satoshi sought to prevent by counting votes on CPU’s, not IP, happened: owners of great computational capacities had a chance to undermine the system and dictate their consensus conditions to the majority of market players.

It’s hard to tell what Nakamoto was thinking when he or she assumed that a system that votes with hashrate would be better protected against minority takeover than an IP-based voting system. It might mean that bitcoin was conceived as a short-term social experiment and the creator didn’t really believe in its success. It’s quite probable that Satoshi would have offered something else had he or she known that bitcoin price upsurge would make it profitable to buy special ASIC processing units and create giant mining farms.

Byzantine Tycoons

The bitcoin community has an inner caste of miner tycoons who control the biggest pools. They formed a core of the industry’s financial stars. Bitcoin is still a p2p network, yet users don’t have equal votes.

This gap is especially important in the light of recent events. When hard fork is on the table, regular users feel like they’re on the back seat of a car racing in the dark. The network is no longer in their hands; just like the outdated fiat currencies, it is controlled by endless intrigues of stakeholders.

The joint statement from a number of leading exchanges about the possible hard fork emphasizes the need for consensus in the community. It’s quite obvious that, unfortunately, social consensus isn’t of much importance in the formula. The machine consensus is what’s it all about. If Bitcoin Unlimited capacities outsail those of Bitcoin Core, the BU fork will become the main blockchain whathever the exchanges call it. Those are the rules written by Satoshi Nakamoto.

The truth is that bitcoin’s consensus mechanism is no longer good for maintaining the status quo, let alone everything else.

Proof-of-Stake

The problems with PoW became evident to some community members long before the sickness has manifested itself. Back in 2011, a concept for a different consensus algorithm, known as Proof-of-Stake, was developed. Its basic principle relies on voting with available resources (coins frozen in a wallet) instead of using external resources like electricity and computational power.
Any participant may be randomly selected as a creator of a new block, i.e. a validator, and this chance is proportional to the coin supply in a user’s wallet. Validator don’t have any additional rewards for creating blocks, though, on the other hand, they don’t lose or spend anything to create a block.

PoS certainly saves power and excludes a PoW-like race of arms. However, because it doesn’t spend any resources, it entails a different set of problems.

First, the more coins a user has, the more chances he or she gets to compromise the entire system. It’s much easier to conduct a double-spend attack or a deep fork (creating an alternative chain rooted in a past block) in PoS.

Second, the Nakamoto vision of consensus may be compromised by holders of bug resources (so-called whales). One may say that PoS has a greater centralization trend: if the major shareholders collude, no one will ever overcome it.

While PoW requires miners to spend lots of resources, thus making it unprofitable to process several forks, in PoS one may use all forks simultaneously. Therefore, a PoS-based cryptocurrency hard fork will entail much greater chaos. That’s possibly one of the reasons why ill-fated Ethereum takes its time to keep its promise and switch to PoS.

Cabinet of Wonder

The quest for a perfect consensus algorithm is still on, with hybrid solution becoming more popular. A classic hybrid is a blockchain where blocks are added by using PoW and PoS alternatively.

A more exquisite hybrid, Proof-of-Activity, combines PoS and PoW: miners solve a cryptopuzzle, however, a block has to be signed by several random users prior to entering the main chain. If a validator is unavailable, the next found block and a group of validators are selected instead.

There are other hybrid configurations embracing both PoS and PoW. For instance, Dash uses PoW for mining and various iterations of PoS to instantly validate transactions with masternodes.

However, the experience has proven that hybrid algorithms inherit the problems of both their ancestors.

A rather peculiar Proof-of-Burn offers you to destroy your own coins as a payment for the right of being a miner. Therefore, the more coins you burn, the higher your chances to find a new block. It is a sort of a refined version of PoW: instead of burning a vast amount of very real electricity, a user removes all intermediaries and literally burns the money. The mechanism is way greener than PoW yet it retains the ancestral problems and doesn’t work for new blockchain systems where issuance of new coins is booming.

There are some peculiar protocols employed by some DAO’s. Mining in those ecosystems is replaced with activity that sustains the organization’s operations. For instance, Storj, a decentralized cloud storage, uses Proof-of-Storage using user disk space as a resource that the system may freely use for its needs.

Those mechanisms work great in DAO ecosystems, yet they’re hardly applicable in global projects like bitcoin.

The Advent

Considering the existing institutionalization and marginalization trends that tear the entire industry apart, one might expect a philosophic and ideological schism to strike developers and users of cryptotechnologies.

Still, the system strives for balancing itself by getting rid of the most radical elements. The most centralization-oriented players found out they actually didn’t need any blockchain. Radical centrifugal forces find themselves beyond capital flows and leave user radars (sometimes entering law enforcement radars at the same time).

The modern-day trend stipulates that new consensus algorithms focus on improving system scalability and accelerating transactions instead of seeking a new solution for Satoshi’s plan where one user has one vote.

A consensus algorithm that heeds earlier problems in terms of social and machine consensus will become a digital messiah whose advent seems so much needed.

Konstantin Golubev

Source