Ethereum: Technical Achievements in 2021 & a Wish List for 2022
Summary: In 2021, Ethereum made major strides on the road to sustainability through advances to accomodate for sharding, proof of stake, Layer 2 scalability, and state growth management. In 2022, a focus on improved wallets, user experience, and privacy solutions will improve the ecosystem even more.
The Ethereum We Have Today
The Ethereum that launched in 2014 is worlds different from the Ethereum we have today. 2021 was a pivotal year for Ethereum that has been in the works since its inception.
Despite the over 300M people that have bought cryptocurrencies through centralized exchanges with US dollars or other fiat currencies, there remains less than 5M users in decentralized finance (DeFi), a category of financial applications on public blockchains which incorporates a redesign of financial systems. You can read more about a decentralized exchange example here. For NFT’s, the numbers are even less, standing at less than 900,000 users in the largest NFT marketplace, OpenSea (Dune Analytics). This conclusively means that cryptocurrency remains largely speculative by over 98% of its holders.
Thoughtful talent aggregation in the cryptocurrency industry is increasingly apparent. Considering the over $30B in investments in Web3 via venture capital, the trend is not wholly unnoticed; however, the edges of innovation in Ethereum are largely left unexplored by incumbents that aim to be “fast followers.”
Unlike the onset of the internet, which invited participation and network effects to achieve success, achievement in a decentralized Web3 will likely be defined far more emphatically by thoughtful builders, creatives, and system designers, rather than the followers that seek to rapidly implement. (See thoughts for creatives here.)
Three major technical leaps were achieved by the Ethereum community in 2021.
1. The launch of the Beacon Chain.
2. The launch of Layer 2 systems.
3. State growth management & storage access.
Launch of the Beacon Chain
The Beacon Chain deployed in December 2020, but it’s operation and continuation throughout 2021 is notable progress for Ethereum. The Beacon Chain initiates the transition from proof of work to proof of stake in Ethereum and allows for the Ethereum blockchain to ‘shard’ into 64 separate chains.
Running alongside the Ethereum blockchain, the Beacon Chain currently only supports itself, but it will support the coordination of staking between these 64 different shards, ensuring the chance of a takeover by validators (or miners) is less than one in a trillion, keeping the network secure. This means that smart contracts and decentralized applications (such as DeFi) are not ‘stored’ on the beacon chain. The main Etheruem blockchain will become one of the shards itself, still supporting smart contracts and transactions. The other shards will support data only, allowing Ethereum to concentrate on rollup centric scaling, discussed in the sections on Layer 2 below.
As mentioned above, the Beacon Chain also supports the transition to proof of stake, a topic that remains surprisingly underreported. Ethereum currently uses proof of work to secure the network and choose the ‘winning’ miner to produce the next block. In proof of work, miners participate in a random cryptographic puzzle to ‘guess’ an answer — the first one to guess the correct answer wins the block and gets to produce the block, winning the block rewards (or newly minted Ether plus transaction fees) in the process. Producing the block is less than 1% of the energy that is consumed in this process while 99% of the energy goes into this proof of work ‘puzzle.’
In proof of stake, that ‘puzzle’ is replaced with another method. Now, miners, or validators, will lock a certain amount of collateral, Ether cryptocurrency, in a smart contract signaling their intention to patriciate as a validator in the network that wants to produce blocks. The winner of the block is randomly chosen from the those participating in proof of stake, rather than a ‘winner’ in the proof of work ‘puzzle’.
Most, if not all, newer chains use this proof of stake consensus — Solana, Polkadot, Avalanche, Near, etc. The proportional amount of cryptocurrency used in proof of stake differs wildly, which will likely influence the economics of the chains in the future in different, still yet unknown capacities. For example, in Solana, over 77% of SOL is staked, while in Ethereum 7.8% of the total Ether is staked.
In contrast, Aleo is a novel blockchain that attains privacy with smart contracts from the onset through zero knowledge cryptography, unlike the chains already listed; however, this design relies on a method called “proof of succinct work”, which verifies cryptographic proofs instead of replaying transactions — thereby achieving efficiencies that are notable over conventional proof of work and will likely prove more advantageous as the growth in zero knowledge cryptography hardware continues. (If a Samsung cell phone has a quantum random number generator these days, it is not hard to imagine cell phones housing zero knowledge hardware in the future.)
Launch of Layer 2 Systems for Scaling
In the blockchain trilemma, coined by Vitalik Buterin, blockchains seek to optimize three things: decentralization, security and scalability. It is important to consider scaling to mean the ability of a blockchain to accommodate more users and applications in a decentralized and secure design, not just in speed alone.
Although critical to a smoother user experience, the sole metric of transactions per second is currently best served by a centralized database, rather than a core blockchain. In the future, speed will continue to be addressed by zero knowledge rollups transaction verification over transaction replay in blockchains, which I will go into more detail later in this piece.
If you have not yet dived into the world of “Layer 2” systems on Ethereum, 2022 will likely leave you unable to avoid it. Layer 2 systems are those that use various cryptographic proof systems that leverage the security of the Ethereum blockchain to enable scaling. These scaling systems were designed for a myriad of reasons: lesser transaction fees, managing state growth, implementing a spectrum of privacy, etc. Layer 2 systems are built “on top” of the Ethereum blockchain.
Note if a system allows deposits or withdraws to the Ethereum blockchain yet does not use Ethereum as a security layer, it is not inherently a Layer 2 system. Examples include side chains, which do not use the security of the Ethereum main network to secure their activity. Therefore, be conscious that everything that allows deposits and withdraws to and from Ethereum is not necessarily a “Layer 2” system.
More generally, in a Layer 2 system, users transfer their tokens from the Ethereum main network to the Layer 2 system via a “bridge” — users see this as a user interface, and what happens under the hood is effectively a deposit into a smart contract. Once in the Layer 2 system, fees are drastically reduced from that of the main Ethereum network and users can explore any application built on the Layer 2 system. At the time of writing, the most competitive Layer 2 system drops fees to less 1% of those on Ethereum main network (Source).
The initial deposit into Layer 2 can still incur significant transaction fees. For Layer 2 ecosystems to thrive, exchanges need to support direct deposits and withdraws to these systems. In this way, a user can avoid withdrawing their tokens to the main Ethereum network from their exchange account, and then depositing them in a Layer 2 system — thereby saving a transaction. Some exchanges are beginning to support this with Binance’s support of withdraws to Arbitrum, and Coinbase’s announced intention to support withdraws to Polygon.
Optimistic Rollups & Zero Knowledge Roll Ups
There are two key Layer 2 systems at the moment: optimistic rollups and zero knowledge rollups. In optimistic rollups, a system of ‘sequencer’ nodes submits batches of transactions to the Ethereum network. Verifier nodes watch the submissions of the sequencer nodes to verify to transactions have been executed correctly. If there is a discrepancy, the verifier submits a ‘fraud proof’ to begin a period of dispute to rectify the transactions. Hence why the system is ‘optimistic’ — optimistically, if there is an error, the verifier nodes will catch it.
Admittedly, another potential drawback of this system is the withdraw times. If you would like to withdraw your tokens within optimistic rollups back to the Ethereum main network, you must wait for a ‘challenge period’ of one week to ensure no one runs away from the system with extra tokens erroneously, being that the system uses verifiers and fraud proofs to ensure integrity. This average one-week challenge period is not without solution. It is estimated that service models will arise out of lenders who are willing to let people receive their withdrawn tokens as an immediate loan for a fee to enable faster withdraw times.
In the market, five different optimistic rollup systems have launched with Arbitrum as a leader in gaining 42% of the market share value wise. Across these systems several decentralized applications are already operating including most decentralized finance use cases.
Zero knowledge rollups (zk rollups) are similar to the methodology used in optimistic rollups — submission of bundled transaction information to the Ethereum mainnet via a sequencer or validator; however, there is one key fundamental difference: validity proofs. Instead of the reliance on fraud proofs to watch for erroneous transactions, zk rollups use validity proofs that allow for a guarantee of the integrity of transactions in the system. Using these validity proofs also allows for ‘instant’ withdraws from zk rollups — or rather withdraw time would be only limited by Ethereum block time (~15 seconds), not a challenge period.
The zero-knowledge terminology is borrowed from a privacy preserving technique called zero knowledge proofs, which employs a combination of probabilistically checkable proofs and forms of encryption. Zero knowledge is a world of its own that may even eclipse the excitement in blockchain at a later date. You can read more information here and here. The largest zk rollup systems remain as singular applications — dydx and Loopring. However, with the launch of Starknet in November and others impending for the next year, an ecosystem of zkrollup applications will continue to grow.
For rollups, it is crucial that the transaction batching that occurs back to the main network is optimized. Imagine millions of rollup systems — that would reintroduce similar transaction fee challenges. Hence, a current Ethereum Improvement Proposal (EIP) aims to reduce rollup transaction fees by ~5x. You can read more about the proposal and other parts of the roadmap here.
While it can be argued that optimistic and zero knowledge rollup systems are in an arms race for users and applications as they build out more features, both remain in experimental releases with centralized controls in place until models are thoroughly tested.
State Growth Management & Storage Access
State growth refers the fact that as the blockchain continues to produce more blocks, anyone who is hosting a node must continue to allocate more storage to accommodate for more data. This problem can sound innocuous, but it is one of the larger threats to blockchains that remain quizzically under discussed. Full archival nodes are very important in Ethereum or any blockchain — these are nodes in which the full history of the blockchain is stored. If there was ever a catastrophic failure in any blockchain, full nodes would be pivotal to recreating and rectifying the system. Full nodes are also required to retrieve historical data. Note that all nodes in a blockchain do not employ the full history of the blockchain as different nodes use different efficiency driven designs.
By the beginning of 2021, the Ethereum blockchain ‘state’ had grown to a whopping 6TB. With the exponential growth in Ethereum, it is critical that state growth management be given serious priority. This year, the Erigon project condensed the requirement to run an archival node down to 1.6TB — with very little press but a 75% reduction in storage requirements is pretty staggering.
This concern for state growth even has some public blockchains targeting succinct state size as a feature, such as Mina. The Mina blockchain is a constant 22 kb.
State growth compression is one thing; however, of equal importance is that ability of applications and users to access data quickly.
Before Google, searching the internet required users to search the unorganized internet for the information they wanted, which was slow and inefficient. Google created a method of storing and indexing information such that users could quickly find what they wanted. In a similar way, a project called the Graph has created a decentralized method of storing and indexing the Ethereum blockchain information through subgraphs. The Graph ecosystem employs several different roles — nodes of indexers, curators, consumers, etc., aiming to create a truly decentralized way of allowing for fast access to valuable blockchain data.
In mid-July of 2021, the Graph had over 25B monthly queries. The Graph aims to support other blockchains in the future such as Near, and the Graph is also adding capabilities for Layer 2 solutions.
Archival nodes and data retrieval are key aspects of robust public blockchain ecosystem. Ethereum projects continue to demonstrate iterative progress in both arenas.
A 2022 Wish List
There are more achievements in Ethereum in addition to the above; however, in working in this category, I have identified a key ‘wish list’ to continue to enable the ecosystem to thrive. This wish list is not without work — many teams are working on these solutions. However, it is key to bring some ideas to the forefront for focus.
1. Safer, functional wallets and accounts for individuals and enterprise.
2. More design consideration.
3. Increased privacy.
Wallets & Accounts
As an individual user, there is nothing easier in public blockchain than spinning up a self hosted wallet, but there is nothing more difficult in public blockchain than keeping it safe.
In contrast, custodial wallets in exchanges like Coinbase or Gemini are key to allowing users to exchange fiat currencies for cryptocurrencies. While these custodial wallets provide a layer of security in protecting cryptocurrency ownership, these wallets do not currently allow for participation in decentralized applications. Users cannot use their custodial wallets to donate to open source projects as in Gitcoin or participate in decentralized lending, as in Compound. There is a specific reason for this, and that entails a platform’s ability to let you sign whatever type of transaction you would like to send to the blockchain. In an exchange, you can use ‘send’ and ‘receive’ transactions (hopefully) liberally at their discretion of whatever tokens are supported; however, if you want to build reputation for a DAO (decentralized autonomous organization), deploy a smart contract, or vote in a DAO, you will need a self-hosted wallet to sign that transaction.
A self-hosted wallet is one in which you are in complete control of the private keys so that you can sign any transaction. However, if you lose this private or a hacker steals it, your wallet is no longer in your control, leading to loss of funds. Metamask is a service that allows users to create self-custodial wallets to interact with decentralized applications. Metamask crossed 10M downloads last year; however, is it rational to expect this kind of growth to continue and for users to continue to ‘proceed at your own risk’?
Without finding a way to adequately secure self-custodial wallets, decentralized application growth will remain limited. Self-custodial wallet teams have created methods that allow for social recovery of self-custodial wallet private keys. You share pieces of your password with a few friends that can help you piece it together, if lost. However, uncoordinated social recovery is placing a big emphasis on individuals’ desire, time, and capacity to organize 3+ friends for wallet recovery. I remain unconvinced this will work at scale but would be happy to be proven wrong.
Multi-signature wallets are those that allow for multiple accounts to control the actions of wallet do de-risk activity. In multi-signature wallets, m of n parties must sign a transaction to execute. For example, you may have three different accounts on a multi-signature wallet, and two of them may be required to sign before the transaction is processed.
However, for multi-signature wallets to be rational, keys must be stored in separate locations. For example, if all private keys are stored in the same server and a hacker gets into that location, stealing three keys will be nearly just as easy as stealing one. So, institutions will need a clear path for accommodating this, which although technically feasible may include legal and security concerns for which many are unprepared. Also, multi-signature wallets for individuals induce the same challenges of motivation, time and learning curves as social recovery.
Besides stronger wallet security, what we would stand to benefit from is self-custodial wallet insurance. For comparison, credit card fraud losses totaled $28.65B in 2019. Crypto hacks led to $9.8B in losses in 2021. However, we do not have the data on the amount of self-custodial wallet hacks. The $9.8B refers to code or system hacks. Even if unable to quantify the self-custodial wallet hack problem, the larger problem is being able to prove that someone hacked a self-custodial wallet and stole a key. Who is to say that you didn’t just swipe the funds yourself and are now trying to claim insurance? However, does there need to be proof of fraud under a certain threshold? Credit card fraud is not investigated under $10,000, but crypto carries much different implications than payment solely for goods and services as in credit cards.
Throughout the last year, bullish believers announced, “the institutions are coming.” Well, until there is a way for an institution to meaningfully interact with decentralized applications in a secure way, these potential institutions will unfortunately only join the 98% of the population that is largely speculating unless they have a leadership team that understands the vision and trajectory of public blockchains as well as supports the exploration of truly participating and building a future in it.
Account abstraction may also pave the way to better solutions as well, allowing for smart contract functionality to enhance wallets from the start. Multi-signature wallets are effectively smart contracts, but the accounts that operate them must be managed by what are called ‘externally owned accounts’ or a simple private/public key pair that does not house any code as a smart contract does. Account abstraction is about making all accounts to use smart contract logic from the onset. This post by Nethermind, a prominent team contributing to the Ethereum ecosystem in several significant ways in zero knowledge as well, goes into more detail.
All this being said, improved wallets are critical.
More Design Consideration
Statistically, every time you require a user to click on something you lose something on the order of 95% of your potential customers. A basic swap on a decentralized exchange (DEX) may only require 4 or 5 clicks, and many DEX’s have mastered an optimized user flow. However, no matter how many clicks, most potential customers are lost at some point along the way. When moving through a user flow, even micro annoyances, such as the need to scroll down to click a button or unloaded image, may lead to a user jumping out of a flow. Ruthless web application design scrutiny is important for bringing the ecosystem out of an intimidating niche to an approachable environment.
For example, digital signatures are now ominous territory. A few hacks occurred last year through hackers distorting the transactions that users thought they were signing — all of the sudden instead of clicking to claim rewards hackers convinced users to sign away their tokens! This happened because users see a random jumble of code when they sign certain messages in Ethereum that do not correlate to a human readable user interface. Effectively, in some very specific advanced cases, users did not know what they were signing. However, this kind of message signature as a type of transaction saves on transaction fees with only the best of intentions from the developer teams. How can we prevent this in the future? It is imperative that trustlessness is not only in the core protocol of Ethereum but also in its decentralized application user interface design.
In the same way the iPhone made users feel like a genius who entered a world that anticipated their issues and made things as clear as possible, so every decentralized application needs to be designed. Things like accessibility are more important every day — would a blind or vision impaired person be able to use any decentralized application right now? These questions are important because these groups of people can now more easily navigate our legacy financial system thanks to design. The experience is not perfect by any means — but what if Web3 made it better?
User interface/user experience design is an amazing field that has barely scratched the surface of decentralized applications.
More Privacy
With the exception of private transactions on Zcash, Monero, and some very specific applications on Ethereum, public blockchains display all information publicly. At this point, given that 99.98% Americans can be identified with 15 data points, even “pseudo-anonymity” seems like a falsehood in the future. Pseduo-anonymity of public blockchains means that unless an observer knows your public wallet address, they cannot identify your activity on a public blockchain. However, I am stipulating in the future, if we all start using public blockchains without privacy, observers will be able to identify us by our transactions.
How does blockchain achieve privacy in a system that is built on consensus? Privacy is inherently challenging in blockchains because validators, or miners, have to validate which transactions are credible to process — hence, they are able to prove to the transaction is valid. In blockchains, this is done by replaying transactions, or running them again. How can you do that if you want transactions to be private?
Zero knowledge proof cryptography allows allow one user, a prover, to prove to another, a verifier, that a computation has been performed correctly without revealing inputs. For example, you could prove to someone you are of legal voting age, without revealing your actual age. This is the type of cryptography on which systems like Zcash and Monero are based.
Zero knowledge proof implementation is no panacea for all privacy concerns, but it is of major design consideration for privacy on Ethereum — being that verifiers can be codified in smart contracts, allowing users to act as provers of their activity while remaining private.
Suffice it to say, several groups are working on privacy models in Ethereum and elsewhere in public blockchain. Aztec is a group working to bring zero knowledge privacy to decentralized finance. Starkware has proposed a Layer 3 system which incorporates privacy over Layer 2 systems. In the Polygon network, privacy designs are also in development with zero knowledge. The zero knowledge challenge is hard — generating proofs can be time consuming and standardization and interoperability is still developing. If you’d like to learn more, I would suggest an initial listen to Howard Wu here.
Recall that zero knowledge proofs are also used for scaling in Layer 2 systems — these designs do not inherently incorporate privacy. Layer 2 leverages the ‘proof’ part of zero knowledge, the privacy aspect of zero knowledge is only addressed in zero knowledge applications specifically targeting privacy.
Privacy is essential for individuals but also for business. The privacy of Ethereum applications is essential for the continued growth of network.
The Year Ahead
2022 will be another pivotal year for the core technology of Ethereum to evolve. Years of evolution brought an ecosystem with unique use cases, such as decentralized finance and decentralized autonomous organizations (DAO’s). Diligent attention to strong core infrastructure rooted in decentralized not as an ideology but as a strength will remain a differentiator. As user experience betters and privacy is achieved, growth should follow — though not guaranteed without overcoming other challenges in regulations and clear legal frameworks.
The developers, designers, users, community managers, lawyers, government policy advocates, etc. are owed both a celebration of achievement for all that has been accomplished and a bout of encouragement for all the work left to do in charting the decentralized course of Web3. Happy New Year everyone!