Algorand and the Blockchain Trilemma

Vitalik Buterin described the “Blockchain Trilemma” as the challenges faced by developers in building an infrastructure that is Decentralized, Secure and Scalable at the same time. It is widely believed that one of the three concepts must be sacrificed in favor of the other two. More often than not, it results in a trade-off between the three.

What’s more, too much focus on one aspect can have unintended consequences. For example, as we have previously argued, Bitcoin’s resource-heavy Proof of Work (POW) protocol, which is based on an incentive system that rewards miners for securing the network through the use of computer processing power, has not managed to scale nearly enough to be used as a medium of exchange and is now centralized in many aspects. Mining has become so resource-intensive that it is only done by large operations while the only reward for users or early adopters is price appreciation. This encourages hoarding rather than usage. To add to the problem, the resulting centralization in mining and ownership can in turn have security implications as it creates single points of failure.

Ethereum, which was also reliant on a less resource-intensive POW protocol suffered from similar consequences due to the added complexities Smart Contracts bring to both scalability and security. As we write, Ethereum is switching to a Proof of Stake (POS) protocol which both rewards and punishes those securing the network financially and according to their stake. While this will solve scalability as “miners” will be replaced by “stakers” the jury is still out when it comes to decentralization. There will certainly be large staking operations and some like Jack Dorsey argue that the development is so complex that most of the ecosystem belongs to Venture Capitalist firms.

Here comes in Silvio Micali, Algorand and “Pure Proof of Stake”. It is worth mentioning that according to Wikipedia, he is the winner of the Turing Award for his work in cryptography and one of the co-inventors of Zero- Knowledge Proofs which are at the center of Ethereum 2.0 scalability solutions.

Algorand uses game theory in a completely different way. To paraphrase Silvio Micali, there are “no incentives”. Every token holder is rewarded with ALGOs just to hold ALGOs and there are no extra rewards for mining or staking. Algorand chooses to incentivize its users as opposed to only rewarding those who generate the next block in the chain (miners). Mining is made extremely simple so that miners are satisfied with their extra work only by the gratification of benefiting the system as a whole. The incentives between miners and users are essentially reversed to those used in Bitcoin. Why be a miner? Well for the greater good…and because you are staking your time and efforts into building tomorrow’s infrastructure. In fact, that is exactly what early developers and computer scientists in Bitcoin and Ethereum have done, and they have been handsomely rewarded by capital appreciation only.

The other novel approach is on how miners are selected to mine the next block. Randomness is heavily used in Bitcoin & Ethereum through SHA 256 algorithms and Hashing. Mining is however rewarded for hard work which is time-consuming as miners compete to solve cryptographic puzzles. Not only does it take 10 minutes to mine one block, but multiple blocks can be mined at the same time which implies that a transaction is only really confirmed after 6 blocks…Imagine waiting 60 minutes after paying your bill…Add to that the electricity consumed to mine that transaction…Not exactly your Apple Pay…

In Algorand, as described by Professor Micali, one user is first randomly selected to construct and propagate a block while a very large set of other users are also randomly selected to agree or disagree on that proposed block through a revisited form of the Byzantine Agreement. The main assumption here is that most of the society is honest and therefore that most of the money is in honest hands. The layers of randomness used at the proposal stage and at the agreement stage ensure that the block and its contents are correct by reducing the odds of bad actors colluding or bribing honest ones to nil.

Committee members select themselves by running an individual cryptographic lottery where winners get the right to enter the Byzantine Agreement. The chance of winning the lottery is proportional to the amount of money one has in the system. Because the self-selecting cryptographic lotteries are run internally it makes it impossible for someone to try and corrupt a committee member.

The computations are made simple so that heavy system requirements or potential economies of scale do not result in centralization.

The result is a blockchain with no forks as only one block is presented at a time which means transactions are final and do not need further confirmations. They are both fast and cheap for the user and society as a whole. It is carbon neutral.

Transactions consist of 4 steps where a different committee is selected every time. The Byzantine Agreement is only made of simple and short messages which result in transactions being quasi-immediate.

More importantly, the fact that committee members are randomly selected at every step of the process and according to their stakes makes for a protocol that should be less resilient to change. Unlike in Bitcoin, no single entity should be able to repeatedly block governance proposals according to their self-interests. This should be a key success factor as more and more complex functionalities, smart contracts, and other applications are rapidly added to the blockchain.

Algorand certainly is an elegant solution to the Blockchain Trilemma. While it remains to be seen whether its novel incentive design is enough to attract sufficient users and talents to compete with the likes of Bitcoin or Ethereum and their First Mover Advantage, Silvio Micali and his team might have created the infrastructure of tomorrow.

First published here