David's Take: 5 Big Questions Facing Crypto's Future
Ethereum is now eight years old, and I've been around for 6 of those years.
Never before has the path forward been more clear – we've come a long way since we thought writing tweets on the L1 would change the world!
We now know what decentralized crypto-networks are good for and how to scale them. There's so much valuable juice left to squeeze from our current understanding of crypto-networks. It's exciting!
Nevertheless, there are plenty of unanswered questions about what the endgame for crypto looks like. While the fog of war is slowly receding, it is still very present.
Making capital allocation decisions, both personally and in my VC duties, depends on the answers to some of the big questions left in this space, of which there are plenty. All the VCs and builders in crypto are trying to answer the big questions more precisely and quickly than their competition.
I've taken time to identify some of the biggest questions left in the space and how I think about them. So, Bankless Nation, here are five big unanswered questions about the future of crypto.
- Many Superchains or one Uberchain?
- Where is value captured in the Rollup Stack?
- Where does LST equilibrium lie?
- Does Solana get gobbled?
- How do we get price discovery onchain?
This post is a monster, filled with an awful lot of questions – so strap in.
⚠️ Disclosures: I am an investor in a handful of projects mentioned or alluded to below, including Optimism. You can see my full, up-to-date list of disclosures and those from Bankless Ventures here.
#1: Many Superchains or one Uberchain?
We know how Ethereum will scale.
Rollups have extrapolated Ethereum's WWIII-resistant blockspace into an abundance of L2 blockspace. In 2020 and 2021, we talked about this in theory, and in 2022 and 2023, we've seen it go into production.
But there are so many teams executing on this same vision in their own unique ways. The Optimism Superchain! Arbitrum Orbits! zkSync's ZK Stack! Polygon Supernets! Eclipse! There are sooo many different ways to build an L2!
Each represents a strategy for extending Ethereum L1 blockspace into the furthest reaches of the internet. My perceived endgame for crypto is that a blockchain finds its way into every corner of the internet, and Ethereum is producing various blockchain genes, each specializing uniquely, to fill the chainless voids of the internet.
But questions remain:
- Do we need so many different rollup standards? Or does just one framework work? Does my Empire Model for Blockchains naturally extend into an "Empire Model for L2s on Ethereum?"
- Or does Ethereum's Rollup-Centric Roadmap inherently lower the barrier to entry for alternative rollup standards, producing a more pluralist equilibrium of rollup strategies?
- Can app-specific rollups (roll-apps) justify themselves economically?
- Or is it likely that the economics converge all applications to live on a few dominant rollups?
- Will economic and composability incentives force converge upon a single monolithic Ethereum rollup, as illustrated as one of the outcomes in Vitalik's Endgame?
- What properties can we predict this theoretical winner would have today?
Rollups have costs, and only some use cases can justify the costs of establishing and maintaining a rollup network based on their economics alone. These applications will have to find a home on a more generalized landscape, and generalized rollups will compete to provide the largest amount of real estate for the cheapest cost.
This argument captures some composability tailwinds as well. When more applications live on the same chain, optionality increases. The whole becomes greater than the sum of the parts, and the more parts, the greater the whole. Humans naturally migrate to cities, and rollup economics are the same. Who can build the biggest city?
As technology develops, it becomes cheaper! The fixed costs of rollup deployment will decrease over time as the tech matures.
While having to contend with infinite rollups seems untenable, additional areas of research and innovation can help contain this. Homogenous blockspace, abstraction layers, cross-chain execution, contract calls, shared sequencing, and offchain intents will all contribute to corralling the chaos of 10,000 chains.
Having every application live together on one chain is admirable, but there are fundamental limits to scale for monolithic chains. While horizontal scaling via many rollups seems chaotic, plenty of research and development is left to squeeze on multiple fronts.
The idea of the Superchain future for Ethereum is compelling, but it doesn't exactly solve L2 composability.
Ethereum's rollup-centric roadmap produces a path towards infinite scaling by simply allowing chains to be deployed as needed to Ethereum. If one L2 gets congested, just spin up another! But this strategy presents new problems for Ethereum that the Solana community would just love to tell you about – mainly composability issues.
Rollup SDKs like the OP Stack help solve this problem. Shared standards and homogenous blockspace are a huge first step to recomposing different chains back together into one single execution layer, but this time with infinite scale. Add in some shared execution, cross-chain contract calls, some UI abstraction, that magical '???' step, and boom, we have an infinite scale for an L2 Superchain.
There's just one problem.
Optimism, Arbitrum, Polygon, and zkSync all want to do this.
A path forward
It's great if the Optimism Superchain is 1,000 different chains perceived as one, but Arbitrum still speaks a different language as Optimism, as do all the rest of the L2 SDKs.
This is why Arbitrum's version of a Superchain doesn't look like Optimism's. Arbitrum is interested in the single unified Uberchain vision, in which universal composability mechanisms tie together any and every chain.
Rather than producing yet another Superchain, Arbitrum is working in the space between these chains. This area of focus is the interoperability layer between the Ethereum L1 and the settlement of a Superchain. Once a Superchain agrees upon its internal state, Arbitrum wants to focus RnD efforts on inter-chain settlement before collective finality is found by posting to the L1.
It's a compelling vision, and if you squint, you'll see that the above two sections are actually the same. There's potentially a future in which many rollups exist, and chain composability technology innovations allow them to blend and blur together into one single Uberchain. Or... those technologies don't work, and the only way to achieve true seamless composability is by just having one single rollup.
So, is the future of Ethereum...
- A handful of different multi-chain economic zones? (Superchains!)
- A single composable Uberchain? (Superchains + cross-chain composability innovations!)
- A single monolithic rollup?
Place your bets! 🎲
#2: Where is value captured in the Rollup Stack?
Each L2 team wants to propagate their Chain-Development Kit across the crypto landscape. There's the OP Stack, ZK Stack, Supernets, Orbits, and surely more to follow.
Why do they make these things? How do L2s benefit from more carbon-copy deployments of their chain? Since forking is permissionless, how do the L2 tokens capture value when forking is free? The process of forking forks out the token. Why would a forked OP-Stack chain willingly pay fees to the OP Collective? What's the incentive?
- Mantle, one of the largest OP Stack forks, forked an older version of the codebase and stated they currently have no plans to join the incoming Optimism Superchain. They're going to keep their sequencer fees for themselves, thank you very much.
- Meanwhile, Base is contributing 15% of its sequencer fees to the Optimism Collective, effectively the OP token. So Base, as an OP-Stack fork, is contributing value to OP, but Mantle isn't.
What gives? What's the difference between these two chains? The answer: Governance. Without governance, we have disparate, chaotic, messy chains-on-chains-on-chains. With governance, we have homogenous blockspace, code reuse, and shared upgradability. With these properties, we have the foundation to begin fusing 10,000+ messy chains into a chainless user experience.
This has been why I've been uniquely compelled by Optimism's strategy and roadmap above all others. Ben, Jing, Karl, and the Optimism team ran through the Ethereum scaling idea maze and arrived at the logical conclusion of governance far before any other team, and have been "taking the hard path" of figuring out decentralized governance from Day 1. All other rollups are competing on technical advantages, but eventually, that will run out, and they'll have to start figuring out their long-term governance strategy once they arrive at that logical conclusion.
Meanwhile, the OP-stack can just absorb the best tech developed by others while working on digging an unforkable moat of governance since, before, any other L2 team was privy to it.
"Why is governance the logical conclusion for rollup competition?" is a vast subject that's out of scope for this article. I'll have to point you toward the very deep Optimism rabbit hole to figure that one out.
This is my thesis for how the L2 space plays out: L2 token value capture is ultimately derived from governance effectiveness. Nonetheless, here are the questions I still have about this:
- Will a generalized, modular framework like the OP stack be able to absorb the best L2 tech in the same way we've seen the Ethereum L1 absorb peripheral technologies?
- How powerful can the incentive to join the Superchain get? This question is a proxy for measuring 'governance effectiveness'; how effective can L2 governance become? Will it become sufficient to cohere a bunch of messy chains together?
- If cross-chain composability innovations don't produce what's needed to create a chainless experience, what other factors can governance leverage to increase L2 token value capture?
But what about Rollup infra providers?
Governance is only half of the L2 value capture equation. Even if L2 SDKs can figure out token value capture, they'll still have to contend with the "RaaS-sized hole in their business model."
If we think there will be lots of L2 rollups, then that means there will need to be infrastructure to host all of these rollups. Think: Cloud, but for rollup infra.
This is why companies like Conduit and Caldera have emerged. They want to host as many rollups as possible to capture some of the fees that rollups produce.
Rollups-as-a-Service like Conduit are in a thumb war with L2 SDKs like the OP-Stack. RaaSs want fees, and the L2s also want fees. Where's the equilibrium?
There are two outcomes I see:
- RaaS Providers want all the fees and will attempt to thwart L2 teams by routing around them.
- RaaS Providers will accept that they are beholden to L2 teams and will simply receive whatever fees they are given.
I'm L2-biased here, so this reasoning could be checked, but here's how I see it. Say the RaaS Provider is maximally greedy and wants #1.
RaaS Provider: "We have all the infra; why do we need to pay taxes to the software when the software is free to fork?"
So, RaaS Providers just take the OP-Stack and help teams deploy OP-Stack chains using their RaaS, and they collect the sequencer fees of all of the chains they operate, and the L2 tokens get NOTHING. They might even unlock some chain composability benefits by helping all their chains shared-sequence with each other.
The problem here is that we return to the governance problem. Producing a plethora of chains is not a sufficient outcome, and while having a single RaaS provider unlocks some shared sequencing benefits, it's not nearly enough to produce a chain-abstracted outcome that the Superchain endgame requires to be successful. If a RaaS wants to attempt to win a thumb war over L2 SDKs, they'll… need to become their own L2 SDK. This means they're entering the Arena of L2 competition, and they'll ultimately find convergence with needing to work on L2 tech innovation, BD, and the big one: governance.
How will RaaSs actually decentralize their L2 tech stack if their significant competitive advantage is that they run physical hardware in physical locations? If any RaaS produces a monopoly by running every L2 chain, all L2s are centralized to that one RaaS hosting center. To decentralize, RaaSs will need to work on the same things that all the other L2s have already worked on for years.
So, this pushes them into #2. RaaSs will be service providers to L2 teams and will have to compete on fees with other competitive RaaSs, as L2 teams use their anti-trust cudgel of token governance to make sure that no one single RaaS produces a monopoly that gives the RaaS outsized power over the L2 SDK.
At least, that's how I see it. So, my question here is: "Am I right?"
#3: Where does LST equilibrium lie?
The spectrum: One single dominant LST versus a handful of LSTs in tension with each other
Here, I am conflicted. I understand the compelling arguments that market forces will force convergence upon a single liquid LST. I am not naive to these factors. But I am also not so nihilistic that I'm willing to "go down easy" when it comes to forces that run up against very firmly held values and beliefs of the core Ethereum community that protect it. Market forces are not the only thing at play here.
Additionally, the dominance of one LST is correlated to the incentive for a subdominant LST to vampire attack it. As one LST gets over its foundations, the desire and power of a vampire attack increases commensurately. This can, at least in the short term, disturb the equilibrium of a dominant LST.
The big questions are:
- How far will the Ethereum community be willing to go to ensure a diversity of LST options?
- How effective are the tools for suppressing single-LST dominance?
- If (when?) a single LST does achieve total monopoly, how much does that erode the core of Ethereum values?
- Does it at all?
In answer to this last question, Mike Nueder wrote a fantastic blog post that defines the idea maze of this particular question.
For me, the value of pluralism is something that deeply attracts me to this space, and I'd like to see more of it wherever possible. Danny Ryan gave the technical argument for why we should uphold that.
If we do ultimately converge at a single dominant LST, how fast does it take to get there? Is slower better? – imo, yes. What guardrails can we construct while we have the time?
We recently saw a denial of the vote for ARB token incentives towards stETH on Arbitrum, mainly due to the concerns about Lido's dominance over Ethereum. If "market forces" were the only thing at play here, this vote should have passed.
#4: Does Solana get gobbled?
Do Solana and Ethereum grow as independent ecosystems, or do the lines between these meld? If they meld, how much gravitational weight does Ethereum have over Solana?
Long-time Bankless listeners will know my thoughts here.
No one has been able to effectively refute my Empire Model for Blockchains, and a bunch of similar theses (Fat Protocol, L1s are Money) are all in alignment with it. L1s are in heated competition for total dominance, and in the fullness of time, one blockchain will ultimately gobble up all the others. 'Tis the nature of open source systems, especially when you add the jet fuel of economic incentives.
Solana has stood out to me versus its competitors. It's not an L1 fork of the EVM, in which any value it creates ultimately just flows back to the Ethereum ecosystem. It's not Cosmos, with no actual settlement layer or enshrined L1 currency. It's not Bitcoin, with all the non-BTC value stripped and deleted.
Solana has its VM, scaling strategy, and L1 asset: SOL. Solana's entire tech stack is not Ethereum, placing it as far away from Ethereum's gravitational pull as possible. This strategy makes sense to me since Ethereum appears to gobble up everything in its sphere. The best chance for survival for any non-Ethereum L1 is to be far away from Ethereum's influence.
Nevertheless, Solana does not live in a vacuum. Eclipse is porting the Solana virtual machine to Ethereum, taking Solana's execution and settling it on the much larger settlement network: Ethereum.
Chris Burniske thinks Solana can retain its autonomy and that Eclipse is a "Solana Embassy" on Ethereum.
I think Eclipse is "Solana tech committing treason" against SOL's value and defecting to join the Ethereum monetary network and settlement layer. Eclipse fulfills the ETH-maxi thesis that all good technology ultimately finds its way to Ethereum, especially when it's just an execution layer, a layer free to detach from a minority settlement layer to join the more global one.
So, what does the future have in store? Can Solana maintain its boundaries? Can Solana's components successfully incentivize itself to remain inside its borders and not engage in network polyamory?
Or, is it more true that no matter how far away an L1 is from Ethereum's gravitational weight, eventually Ethereum will gobble you up too, and the earlier you can defect from the minority network to defect to the majority network, the better off you'll be?
How strong is the incentive to defect to Ethereum?
I think in the future, we'll be able to look back and see evidence that gives points towards both sides. Both sides will be able to say, "We were right." But one side will be more right than the other, and anyone who's listened to Bankless knows where I fall on this debate.
#5: How do we get price discovery onchain?
Some of the most exciting deals we've looked at at Bankless Ventures have been around this question. Having price discovery occur onchain rather than through Binance will produce massive tailwinds for the entire industry. Price discovery represents a balance of power between decentralized and centralized systems, and so far, price discovery is a trophy firmly held by the centralized camp.
If decentralized systems are going to 'win,' we need that trophy. I want to rip it from the cold, dead hands of Binance. Coinbase, you don't get to have it either. Crypto-economic systems are truth machines, but right now, the source of truth about crypto prices isn't derived from the systems that host the assets. We need to go full ouroboros here. Crypto produces the assets and needs to be the oracle for what their prices are. We're the captain now.
At least, that's where we need to be. Various promising mechanisms can help tilt favor toward decentralized systems, but it's unclear how far this will take us. Binance has the benefit of 1-millisecond block-times. No decentralized system will ever be able to keep up with that, and price discovery naturally converges on the most liquid and fastest-updating oracle.
How can we get this onchain?
Promising innovations that tilt the balance of power lie in the arena of Uniswap Hooks and intents. Intents might be the big unlock here. There is an arena of price discovery that occurs in the space between CEX's and DEX's. This nebulous, undefined space is where market-makers and MEV bots make decisions and execute trades, not on any venue in particular. The space-port of Ethereum needs to become conducive to this plane of existence. We need to build infrastructure that supports trade between onchain DEXs and the infinite variety of spaceships zipping around the space between chains and CEXs so that we can encourage them to be more proximate to us and further away from Binance.
So long as our crypto systems live under the shadow of CEX price discovery, we will forever be an inferior version of ourselves. Achieving onchain price discovery will be one of the most significant signals of the maturity and sophistication of our industry.
This is a non-negotiable; we need this. This is not "Can we?" it's "How do we?" Without onchain price discovery, the crypto experiment has failed in some major way.
The mechanism that produces onchain price discovery will surely be one of crypto's most valuable pieces of infrastructure. It doesn't have to be a single silver bullet, either! Uniswap and AMMs are a huge asset in the crypto tool belt for producing onchain price discovery, but more mechanisms like this are needed.
Who's going to build it, and what does it look like?
This article is about questions, not answers.
Many other "questions" about crypto are being asked in 2023; these are just the first five big ones that come to my mind.
- How do Ethereum composability and chain abstraction work in the context of its rollup-centric roadmap?
- Where does value settle in Ethereum's rollup-centric roadmap?
- How significant is the LST threat to Ethereum, and what does its future have in store
- What does the future of Ethereum <> Solana relations look like?
- How do we get price discovery onchain?
These are all very big questions that will require some very big answers. Rather than a single bullet for each question, the answers to each likely come in various strategies, mechanisms, and projects. This is why collaboration and communication between Web3 builders is critical; no one is going to solve these problems on their own.
Of all the things mentioned here, getting price discovery onchain seems like the most complex and most noble cause, as it's something that's going to plague all crypto systems equally, no matter what tribe you find yourself in.
Understanding the idea landscape for how we solve these problems is still something that I'm working on, so I will report back on this once I know more! If you think you know something I don't, hit me up so I can learn!