In my previous article, I introduced you to Algorand and its origins. In this article, I’ll try to explain the motivations behind the creation of Algorand.
Algorand is a permissionless blockchain platform running on Pure Proof of Stake (PPoS) consensus. It is highly scalable and provides immediate transaction finality. Algorand aims to build trusted infrastructure for the borderless economy — an inclusive and transparent system where anyone can build prosperity, no matter where they are.
Motivations: Why was Algorand created?
Algorand was created to realize the goal of a borderless economy. It was designed to create a blockchain that would feature (1) high transaction throughput, (2) strong security, and (3) actual decentralization. Modern blockchains often go through a tough trade-off when it comes to integration of these three parameters. This situation is popularly known as the Blockchain Trilemma, a term coined by Vitalik Buterin i.e. it is very difficult for crypto projects to satisfy the standards for all of the three parameters stated above. Projects inevitably have to prioritize two of these three parameters, ultimately making a compromise with the third.
To put this into perspective, let’s discuss the trilemma with some of the most popular cryptocurrencies out there …
Unquestionably, the most popular of all crypto currencies. It features strong security (SHA-256) but lacks in the areas of scalability and decentralization. Its transaction throughput of 7 TPS has often been a point of critique. Although, off-chain solutions like lightning have been in development for quite some time, so far their integration with the bitcoin ecosystem has been rather slow. On the decentralization front, although there are a LOT of miners mining it, the advent of mining pools has diluted Satoshi’s vision of true decentralization. According to a recent study, about 81% of Bitcoin mining is done in China. Were Chinese mining pools to join hands for some nefarious activity, things would get tricky for Bitcoin.
Similar to Bitcoin, ETH also features strong security (SHA-3). Although it has a higher transaction throughput than Bitcoin, however, at 15 TPS it is still a far cry from acceptable transaction standards. Additionally, it too suffers from poor decentralization as major mining pools mining ETH are centered in China.
When it comes to scalability and security EOS fares well, but when it comes to decentralization, it scores poorly. Having just 21 block producers grossly undermines the spirit of decentralization in blockchains.
It is a highly scalable, secure, and decentralized network. It uses PPoS as the consensus protocol. In the current mainnet, it can execute transactions at a throughput of more than 1000 TPS. To ensure and enforce decentralization it uses the ingenious method of cryptographic sortition; wherein block-proposers, and the committee members who affirm the block proposed by the proposer are randomly selected. The protocol is designed in a way that at every new round, different committee members are chosen randomly to affirm the new block with their digital signatures. This greatly limits the possibility of adversarial attacks on Algorand as no one knows the identity of the users who are engaged in the current round until they have cast their affirmations.
Besides solving the blockchain trilemma, other motivating factors for the creation of Algorand were
(1) inefficacy of the existing blockchain consensus standards, and
(2) the barriers in achieving the idealistic results of the consensus.
To illustrate it, let’s start with Proof-of-Work (POW)…
Introduced by Bitcoin, it was the first of a kind consensus for a permissionless digital-cash. It received, and till this date continues to receive its well-deserved share of accolades for the technological breakthrough in the field of cryptography. However, realistically speaking, it is very difficult to create any other successful POW coin now. Bitcoin is the oldest cryptocurrency, the number of people mining Bitcoin is humongous. It is very difficult for a new project to attract miners on the scale that Bitcoin has collectively done in its decade of existence. The above-said axiom is affirmed from the fact that a number of POW coins were 51% attacked. Some of the most prominent examples include Bitcoin Gold, Bitcoin Cash, ZenCash, Ethereum Classic and many more. An interesting point to note here is that many of these had a valuation of over 100 million USD in market cap, and some even had a valuation of over a BILLION USD when they were attacked.
Another issue with POW consensus is the wastage of Computational and Electrical Resources. POW coins waste lot of computational power. To mine a new block, many miners are engaged in a competition to find the new block. They have to solve a cryptographic puzzle in to gain the block rewards i.e. in the end only one miner will be the winner. But what happens to the rest of the miners who were trying to solve this puzzle? Their computational resources were wasted. In Algorand, there is no such wastage of computational power (At least not to the scale of POW). In every round, a block-proposer is randomly selected out of all the users in the Algorand network to propose the new block. Not every user is engaged in a 2.5minutes-10 minutes competition to solve a cryptographic puzzle and mine the next block. All they have to do is run a cryptographic sortition function which randomly assigns the role of either a block-proposer or a committee member to the users in the Algorand network.
Coming to Electrical resources, POW coins requires specialized, power-hungry hardware (which are quite costly) to mine POW coins. The electricity and hardware costs of these operations prove to be a daunting entry barrier to a miner who wants to mine these currencies. Whereas in Algorand, any user can actively participate in block-proposal and block-propagation operation even with an ordinary computer; thereby precluding the need to buy specialized mining equipment.
Besides resource wastage, POW coins suffer from one more daunting flaw- Transaction Finality. In POW coins, forks are not a rare occurrence. These forks can go up to add 4–6 blocks depending on computational complexity before true consensus on the longest chain finally emerges. Lack of transaction finality forces merchants to wait for these confirmations which usually takes a considerable amount of time (about an hour when it comes to bitcoin).
Algorand, on the other hand, is fork-less. Any transaction that is made on the blockchain is final and is accepted by every node in the Algorand blockchain.
Bonded Proof of Stake (BPOS)
In BPoS blockchains, the network validators have to stake a considerable amount of money for a certain period of time (a month, three months, yearly, etc.) before assuming the role of validators. It is done to deter the validators from engaging in nefarious activities. However, an interesting question to ask here is…What kind of users will be buying a stake? Will it be a user who has a large capital? Or a user with small capital? Naturally, it’s the user who has large capital. A typical user in the real world does not have a significant amount of money to buy a stake big enough which could have a considerable influence in proposing the next block. In a blockchain where you have to lock off a considerable amount of tokens/money for a certain period of time, effectively acts as an entry barrier to people who prefer liquidity over store-of-value. This thereby indirectly restricts the entry of small capital users to take part in the governance of the blockchain network. Having a small set of users doing governance for the majority-economy of the blockchain poses a serious challenge for effective security of a blockchain. In a case where these small set of users are adversaries (or decide to turn adversarial with time) All, they have to do is wait for an opportune time and initiate the attack. Adversaries will lose their security money held in stake, but will walk-off with more gains whilst leaving the blockchain ecosystem of that particular project in shambles.
Delegated Proof of Stake (DPoS)
The biggest issue when it comes to DPoS is…centralization. In DPoS, validators are elected by users to represent them in the blockchain. The number of validators is very small e.g. in the case of EOS there are just 21 block producers. Additionally, these block producers know the identities of each other, which makes it very easy for them to collude and undermine the governance of the blockchain. EOS’s developers themselves have acknowledged the threat for the same.
In Algorand, there are thousands of nodes in each round that verify and digitally sign the upcoming blocks. No one knows the identity of any other user until the voting has been finished, thereby precluding any possibility of collusion.
In the past twenty years, banks have been involved in one controversy after another, be it money laundering, subprime mortgages or any other financial scandals, their money-hungry attitude has led to trust-erosion in the banking institutions by the public. In an interview, Steve Kokinos, CEO of Algorand said “If you look at banks, twenty years ago…there were forty big banks in United States. Today there are four, and those four banks command double the amount of money than the forty banks did 20 years ago.” Steve also went on to say that “about 6% of the world’s GDP, every dollar that every person earns everywhere, slides out the back door in form of transaction fees and other fees, just due to friction in the system. We think it is an enormous opportunity to improve efficiency by orders of magnitude.” Indeed, 6% of the world’s GDP is an enormous market for Algorand to deploy its fintech innovations.
In this article, I tried to explain the reasons behind Algorand’s creation. How Algorand’s creation was a consequence of the inability of existing blockchain projects to collectively attain- scalability, security and decentralization; and also how it was in part due to existing yet unaddressed pain-points in the current global financial infrastructure.
Comments and Discussions are heartily welcome.