A brief history of Ethereums’ near future.
Note: This article is a thought exercise.
1: Gas prices will scale with the value of the transactions and will continue to be expensive and trend upwards.
The explosion of Defi has changed the value proposition of the idea of a “world computer”. In one vision the idea was that Ethereum would be populated by a rainbow of interesting diverse smart contracts deployed by a diverse array of actors all interested in leveraging decentralized compute.
The success of Defi has changed the narrative of Ethereum and brought new users interested in higher-value transactions who are willing to pay higher gas prices as a function of their transactions being more valuable and the speed of their confirmation being more important.
The introduction of the concept of Yield Farming as potentially the “new ICO” may only accelerate the rate at which Defi transactions consume the network.
The effect of a sustained gas price rise will push lower value transactions or non-monetary transactions off the main network (L1) as they become too expensive in Gas to justify.
While it would seem like layer two solutions (L2) would be ideal for these lower-value transactions, the majority of L2 solutions are currently engineered primarily (not exclusively) for “cheapFi”, or Cheaper Defi, and focus on things like high-speed, low-cost token transfers of some sort or another. Even Reddit Tokens I perceive as a kind of “cheapFi” where people are able to earn some sort of “equity share” in the “social capital” of being on Reddit. I am sure there will be ways to build non-Defi applications on Layer 2, but really I expect them to capture the interactions that individually aren’t valuable enough to compete with the gas fees of L1.
I expect that L1 will trend towards becoming a new kind of “interchange banking backbone” for soon to be extremely large and extremely valuable transactions.
2: Application-specific blockchains instead of Layer 2
For large non-Defi projects like CryptoKitties and Aragon, the incentive is to build an entirely new blockchain. The majority of L2 solutions on the market today are not community projects, but rather VC funded companies looking to capture the speculative value of an L2 project gaining market share.
Why should a project like Aragon bring its customer base to an L2 solution, only to have a different group of people capture the upside value of their audience? It makes more sense to build an entirely new blockchain so that value can be captured at both the protocol AND application layer. Hence we see Flow and Aragon chain. (They also built these blockchains to satisfy technical requirements that were important, but I think my point is still valid).
For smaller non-Defi projects, collectible tokens, small DAO’s, etc… It seems hard to imagine what the true value of an L2 solution will be. Technically they enable cheaper transactions with interoperability, but will interoperability with L1 be very important if it’s incredibly expensive? Why would users ever move their digital cat back to L1 if it costs a fortune to do so and when all the other digital cats are on the L2 layer anyway? It seems more likely that if Flow or AragonChain catch on for their respective use cases, they might simply each become the application-specific blockchains with Digital Cats and non-money-DAO’s leaving L1 Ethereum for good.
I don’t yet have an opinion on how ETH2.0 might change this dynamic.
2: MetaTransactions and Wallet on Layer 1
As great of a proponent of MetaTransactions as I have been, right now it seems like their near term future application is limited. Originally MetaTransactions was billed as the solution to user onboarding, for users who don’t have immediate access to Eth to pay for Gas or as a tool to allow users to pay in their native token or with credit cards.
As gas fees rise, customer acquisition cost becomes prohibitively expensive. One thought is to simply charge the user for their gas fees via credit cards, but if fees continue to rise we can quickly arrive at a point where no one is willing to cover the fee: neither the platform, nor the user, nor the wallet.
This again will most likely kill the non-Defi consumer-facing applications on Ethereum. Already popular composable applications such as KittyRace, which allowed you to pit your CryptoKitties against one another in a car race, appear to be gone but which would have been prohibitively expensive to use anyway.
On an infrastructure level, I am coming to think the appropriate layer for Meta transactions to live is at the smart-contract wallet layer. As the value of Defi increases so will the danger of incorporating dependencies into your codebase as a smart contract developer. I think this will lead projects to avoid building dependencies like MetaTransactions into their codebase because it makes their system fragile, and requires additional considerations regarding upgradability (to account for MetaTxn providers going out of business or being compromised, etc..) which itself brings substantial danger.
The simpler solution is to simply recommend their users to use a specific smart-contract wallet, and if they want to subsidize their user’s gas fees, they can simply refund the wallet providers for their user’s transaction fees. This method completely insulates Defi protocols from dependencies on third parties and lets smart contract wallets focus on the security of their users while protocols focus on the security of their protocols.
However, if Defi becomes the primary use case of Ethereum, even this narrative seems to struggle. As the value of transactions goes up and the price of transactions goes up, the less likely applications on Ethereum are to be consumer-facing and the more cautious and experienced the users of Defi will become. How likely is it that users transacting large amounts of value on Ethereum will be comfortable using 3rd party wallet providers, especially smart contract wallet providers? And how likely is it that people willing to pay $100 to construct complex Defi schemes aren’t going to have the ETH to pay the gas?
It seems unlikely that 3rd party wallets have a compelling use case on L1 once the balance of a user’s account exceeds a certain threshold. This doesn’t mean they won’t have users, but it’s possible the vast majority of value is captured by a small number of high-end users who most likely can’t or won’t trust third parties for managing their funds.
3. A shift in Topology: Developers
For a number of years, Ethereum was Defined by smart contracts with limited complexity. Most tokens were relatively simple contracts and most applications were fairly simple to interact with.
The simplicity of working with Ethereum lead to a large number of developers interested in building in the space, and they built a variety of different types of projects and smart contracts.
As the composability of Defi increases, so does it’s complexity, and today interacting with popular smart contracts has become genuinely complicated. This professionalization of the space means that new developers are required to have a deeper understanding of financial primitives and a clearer idea of what they are interested in building. This specificity and complexity will accelerate the trend towards innovation in financial tools and will shift the accessibility of Ethereum away from general interest developers to more skill-specific and experienced developers with an interest in financial tools.
4. A Shift in Topology: Protocol Diversity
The sudden explosive growth of Compounds’ $COMP token implies a coming new opportunity for raising capital in non-traditional ways. What it also suggests is that the delta between the amount of value captured between protocols is also likely to significantly increase. As network effects begin to accumulate in specific protocols it starts to look like a very real possibility that a small number of Defi protocols on Ethereum will end up capturing the majority of value transacted on it. The composability of protocols will only accelerate the accumulation of network effects as base layer components become permanent base level components of new derived products. As liquidity begins to pool substantially in specific protocols/markets and the gas fee’s for aggregation of liquidity increase, L1 Ethereum will exist to support a small number of high-value protocols and transactions, with everything else pushed to L2 and beyond.
My take away:
I have come to the conclusion that in our near future Ethereum will start to look more and more like the backend infrastructure of a new kind of banking system. This is a good thing and means Ethereum is winning. I personally am a little nostalgic, because I think this means the new developers entering the space will be more professionals interested in custom tech solutions solving specific problems around financial innovation. I suspect it means fewer folks building wild experiments for new ways to organize society on Ethereum Layer 1 however.
Finally, I think most, if not all non-Defi, or non “value-first” applications will either leave for other blockchains, use layer two solutions, or potentially cease to exist. This isn’t necessarily a problem though, a natural evolution of technology and markets. For the time being it might make sense to focus resources on Layer 2 solutions and Defi services.