Since Cryptokitties, people have realized how dire Ethereum's need for scale is. Ethereum in its current state is a functioning World Computer but it is very, very slow. A multitude of new projects have been birthed to become "Ethereum-killers" by using novel consensus algorithms to achieve blockchains that can process thousands of transactions per second. These projects include DFINITY, Thunder, Hashgraph, Algorand, and so on.
This has created an arms race to build the first high-throughput smart-contract platform in hopes of stealing market share from Ethereum. These projects have come out guns-blazing with hundreds of millions of dollars in their war chests, world-class cryptographers and computer-scientists on their teams, and the backing of famous investors. Because these projects are all competing in the same space, this has become a literal race to be first-to-market.
Amongst this race to build Blockchain 3.0, Tezos, a "self-amending cryptographic ledger", sits in an awkward position. Instead of touting thousands of transactions per second, it does approximately 50 tps. It does not have privacy features, use zero-knowledge proofs for blockchain compression, nor use cutting-edge cryptography for randomness. How does this project have a shot at competing with the other heavyweights?
The race is not always to the swift
What differentiates Tezos from other projects is its ability to upgrade itself. The architecture of Tezos allows anyone to propose an upgrade to the protocol and the Tezos community can vote on these upgrades to pass them. Ethereum, on the other hand, only permits a small group of insiders, known as editors and core developers, to include EIPs (Ethereum Improvement Proposals) into the main Ethereum protocol.
Diving more deeply, upgrades work through the following mechanism:
- Developers submit a proposal to upgrade the Tezos protocol, including some request for compensation for the work.
- Tezos holders vote on the proposal. If the vote is passed (some majority), the change automatically gets included into the Tezos testnet
- There is a fixed period of time where this change lives on the testnet. If bugs are found, the Tezos holders can recall the change anytime safely.
- Once the period ends, a confirmation vote is held. If the vote is passed, the change is included into the mainnet.
This is interesting because it is effectively a market-driven way of upgrading a protocol. If a developer creates an upgrade to the Tezos protocol and attaches a $100k request for compensation, the Tezos community can value if the upgrade is worth the $100k or not. If they decide it is, the upgrade will swiftly get passed with no bureaucracy or slowdowns. This is different in Ethereum, where the foundation gives grants arbitrarily based on the foundation's discretion. For example, in this grant report, the team implementing Sharding, Prysmatic Labs, is funded with $100k whereas the team researching state channels, L4, is funded with $1.5m. As a part of the Ethereum community, I believe that sharding implementation is much more valuable and pressing than state channel research, but I have no say in this process. Was this because Vitalik is an advisor to L4 and is much closer with the Ethereum Foundation?
In Tezos, I can implement features which I deem valuable by the Tezos community and get paid if my upgrade is really valuable, without knowing anyone in the Tezos foundation whatsoever. This will incentivize a bunch of independent developers and teams to work on different upgrades to the protocol without a central body of coordination (Ethereum Foundation) deciding what needs to be worked on.
This is important for one main reason: Tezos can start out slow. Thunder must launch with extremely high tps on day 1 to be interesting at all. Projects like Coda must launch with blockchain compression working on day 1 to garner any interest. However, this really does not matter at all for Tezos. With a strong enough developer community, Tezos can slowly copy these "killer features" and implement them into the Tezos protocol over time. While Thunder and Coda may seem interesting in 2018, they may completely fall out of relevancy in 2028 especially since they do not have a sound plan on how to keep improving.
Right now, Thunder's consensus algorithm which lets it achieve thousands of tps or DFINITY's random beacon gives them moats over other blockchain projects. My thoughts on this are perfectly articulated by Elon Musk, as he recently said "moats are lame", and "what matters is the pace of innovation. That is the fundamental determinant of competitiveness." In an open-source world, moats are even lamer, and co-ordinating developers to innovate on protocol upgrades are even harder. Because of this, Tezos' architecture which formalizes protocol upgrades has set it up for long-term success.