AMA review: Can the Ethereum Foundation take action against FUDETH?

Compiled by Yangz, Techub News

ETH's recent weak performance has triggered a lot of FUD about Ethereum in the market, and many investors have raised the question "Is Ethereum still viable?" To this end, the Ethereum Foundation (EF) research team held an AMA event on Reddit last night, with topics including the previously hotly debated L1 and L2, ETH's value accumulation, the operation of the Ethereum Foundation, and technical topics such as Blob and Rollup. Techub News has screened and sorted the AMA for your reference.

About L1 and L2

Q: As the L2 solution matures, are there still plans to further expand L1? If so, what methods are being considered?

Justin Drake: The long-term sustainable plan is to leverage SNARKs to scale L1 EVM execution, essentially unlimited.

With live L1 EVM SNARKing, testers can verify low-cost SNARKs instead of re-executing EVM transactions. This will allow us to increase the gas limit by orders of magnitude without increasing the burden on validators. All the heavy lifting of EVM execution will be done outside of consensus, by specialized nodes operated by entities such as searchers, builders, explorers, etc. Users and consensus participants can easily run nodes on their phones or even watches.

In addition to the vertical scaling benefits of significantly increasing the L1 EVM gas limit, there is also the opportunity to leverage EVM precompiles in the EVM to enable arbitrary horizontal scaling, thereby verifying EVM execution in the EVM at a low cost. This precompilation will allow developers to programmatically launch new L1 EVM instances, unlocking a super-powerful version of execution sharding, where the number of shards is unlimited (rather than capped at 64 or 1024 shards), and a single shard is a programmable Rollup, known as a "native Rollup".

Vitalik: L1 expansion strategies currently being actively considered include:

1. Implement proposals that reduce full node load (such as EIP-4444, Verkle trees or hash-based binary trees, and eventually ZK-SNARKing EVM), and once these improvements are in place (or are about to be in place), the gas limit can be increased. The most realistic option in the short term is EIP-4444, as it does not require changes to consensus and only requires some work on the client code, which is relatively orthogonal to other work being done on L1.

2. Further improvements to client execution, once completed, will further increase gas limits. The key areas for improvement here are (i) execution, including virtual machines and precompilation; (ii) state read/write; and (iii) data bandwidth. There are known inefficiencies in all three areas that can be further addressed.

3. Add features to the EVM to speed up specific forms of computation. One of my favorite features is the combination of EVM-MAX and SIMD, which will provide numpy-like extensions to the EVM, making it faster when doing heavy cryptographic processing. This will make applications that rely on cryptography cheaper, which is a great benefit to privacy protocols, while also enabling L2 to be committed to the chain more frequently, reducing deposit/withdrawal times.

Dankrad Feist: Scaling L1 execution is one of the goals, and it is in parallel with building Rollup. The two are not in conflict.

1. Data availability can be scaled almost infinitely - the ultimate limit is people's interest in Ethereum (i.e. how many people are serious about running full nodes and how many people are willing to record all the data).

2. Execution will always be subject to some limitations, and the ultimate limitation is the single-threaded limitation; currently, state access is the direct limitation to scaling L1 execution.

With zkEVM and parallelization, I still think we will see L1 execution capacity expand to 10-1000 times the current level. Rollup will help achieve "world-class scale". The good news is that all of this work is being done in parallel by many teams in a more decentralized way, thanks to a Rollup-centric roadmap.

Q: What does the Ethereum Foundation think of Max resnicks's accusation that L2 is becoming increasingly parasitic on L1? Why isn't L2 decentralized faster?

Justin Drake: I think he's wrong. I've discussed this with him privately. His core premise, which he's stressed many times, is that L2s don't want to decentralize because they lose sorter fees. He's also shared this on Twitter. In both the podcast and on Twitter, he specifically mentioned Coinbase, which makes $200 million in annual revenue through Base. Counterintuitively, L2s are incentivized to decentralize sorting to maximize fees - the exact opposite of what Max is saying.

The term "sorter fees" is ill-conceived and misleading. 100% of Base's revenue comes from execution congestion fees. Base fees are determined by the gas mechanism in EIP-1559, which mimics L1 (see the docs here). The biggest difference is that Base's fees are sent to the Coinbase wallet instead of being burned like L1.

Coinbase makes so much money because the gas demand on Base is greater than the gas target. This is a VM throughput problem, and congestion fees are essentially irrelevant to sorting. If Base uses a decentralized sorter, then Coinbase can still earn this fee. For example, if Base uses L1 validators for sorting and becomes a "based rollup", then Coinbase will still charge execution congestion fees. Congestion fees come from the Base VM Gas target. The sorter only informs users of the L2 congestion fee and plays a superficial role. Value creation comes from the mismatch between the base virtual machine gas target and the supply and demand of block space. (Additional note: Base's transaction sorting is a priority fee auction, first come first served, which means that CEX-DEX arbitrage MEV will be included in Coinbase as sorter fees. Better DEX designs will not give MEV to the sorter, but give it back to LP.)

In my opinion, the only valid use of the term "sorter fee" is in MEV. In MEV, value capture specifically comes from sorting, that is, the strategic positioning of pre-running and post-running transactions. Base's sorting is first come first served, and users send transactions end-to-end encrypted to the Base sorter through a private mempool. Coinbase does not capture MEV, and there is no sorter fee! I don't know which L2 captures MEV now. Base captures MEV at the expense of its users, e.g., swappers will be caught in the middle, or DEX LPs will be more affected by adverse traffic, ultimately leading to a drop in swapper prices. L2 naturally does not want to reduce the quality of execution for users, so L2 does not extract MEV.

In addition, it turns out that sorter fees are not only bad for execution quality within L2, but also bad for composability across L2s. In fact, to extract MEV, some kind of proprietary sorting infrastructure is required, and shared sorters are excluded from the two proprietary sorters. Without shared sorters, the gold standard known as "synchronous composability" is lost. See "Why is synchronization valuable?". The weakening of composability reduces the opportunity for cross-L2 transactions (e.g., transactions from DEX aggregators such as 1inch), which ultimately reduces congestion fees. In order to maximize fees, L2 should maximize congestion fees, which means maximizing composability.

To maximize composability, we need a shared sorter. How do we, as a community, coordinate on a typical shared sorter for all of Ethereum? Two competing L2s (e.g. Arbitrum and Base) can only agree to choose a trusted neutral shared sorter. In my opinion, only a decentralized and permissionless sorter can achieve sufficient trusted neutrality. As some of you may know, I have a stronger argument: in my opinion, the only trusted full Ethereum sorter is Ethereum itself, which does not introduce new brands, new tokens, or new security assumptions.

As for why L2s are not decentralized faster? In fact, sorter decentralization is hard and takes time. L2 currently uses centralized sorters for three main reasons:

1. Security: Even if there are fraud proofs or SNARK vulnerabilities on the mainnet, centralized sorters can prevent attackers from exploiting these vulnerabilities without authorization. Decentralizing the sorter means having multi-validation, formal verification, or some other security mechanism (such as TEE).

2. MEV: Centralized sorters provide a fast encrypted Mempool to prevent MEV withdrawal. Decentralizing the sorter means having other complex encrypted Mempools, such as SUAVE or some other suitable MEV channels.

3. Pre-configuration: Centralized sorters provide a fast user experience. Decentralizing the sorter means achieving low-latency consensus or using cryptoeconomic pre-confirmations, which are actively being developed.

About ETH's value accumulation

Q: What is the logic of ETH's value accumulation in 2024?

Dankrad Feist: Ethereum is building a financial platform, which will be the most neutral platform to date, allowing the issuance and trading of financial assets, and also allowing the creation of new financial products such as derivatives on its basis.

This is a very valuable activity. I think that ETH relies on some kind of fee mechanism to obtain value accumulation from it. In the upgrade roadmap, I think Ethereum L1 will be the intersection between all these subdomains, and many valuable activities will continue to be carried out on it, which will also generate considerable fees. (To achieve this, L1 must be properly expanded)

If this is not the best value accumulation mechanism, there are some interesting alternatives, but I am personally skeptical of these. However, they are not impossible. For example, using data availability fees as the main asset trading medium in the ecosystem and finally using them as collateral (the riskiest) to achieve value accumulation.

Justin Drake: ETH is Money :)

Anders Elowsson: ETH will generate value when Ethereum promotes sustainable economic activities.

By "sustainable", I mean activities that bring utility to participating economic entities and ensure that these activities can continue for a long time. In this case, native ETH will generate value because ETH is a trustless asset in the Ethereum system and is therefore suitable for holding and using as currency. Additionally, payments for economic activity settled on/secured by Ethereum are denominated in ETH and destroyed, effectively distributing value to all ETH holders.

Q: Does the Ethereum Foundation believe that continued appreciation of ETH is important? Why?

Dankrad Feist: The Ethereum Foundation does not comment on this, but as researchers we all have our own opinions. Personally, I think the best approach is to focus on building an ecosystem on Ethereum that can create value, and I think value capture will eventually come naturally. This doesn't mean I haven't thought about it, but it's a big mistake to focus on value capture when the value creation part hasn't been completed.

Justin Drake: I personally believe that the value accumulation of ETH is critical to the success of Ethereum. I don't think ETH can become the settlement layer of the Internet of Value without Ethereum actually becoming the programmable currency of the Internet. Monetary premiums only accrue to the size of a particular asset (think tens of trillions of dollars). This monetary premium needs to underwrite trillions of dollars of decentralized stablecoins ("economic bandwidth"), provide unquestionable security even from nation-states ("economic security"), and attract the attention of all major economic players ("economic salience")

Anders Elowsson: I think quite a few researchers think that ETH's value accumulation is important. One obvious reason is that Ethereum is secured by staked ETH, so ETH's value accumulation ensures Ethereum's economic security. Second, ideally, a currency should maintain its value over the long term, and ETH is the best currency for Ethereum. In a decentralized economy, having a reliable, trustless currency is valuable, so ETH's value accumulation will make Ethereum a better platform. In addition, a large portion of future investments in the Ethereum ecosystem may be held in ETH. This also includes the (not particularly large) treasury of the Ethereum Foundation. Finally, ETH's value accumulation is closely tied to Ethereum's success.

Q: If the end result is a diverse Rollup ecosystem on Ethereum L1, a large number of dapps on L2, and transaction fees below 1 cent, but there is little or no value accumulation in ETH, would EF consider this a successful implementation of the Ethereum roadmap?

Dankrad Feist: When I create a startup, I certainly want to make money, but if it is useful to my customers and I end up with nothing, I will still consider it a success.

Translating this to Ethereum, I do think that if we have a diverse ecosystem that provides interesting applications to the world, I will consider this a success, but if this ecosystem can also make ETH more valuable, it will be even better.

I know a lot of people think that a Rollup-centric roadmap will take away fee income and MEV from Ethereum, and that Rollup could end up parasitizing Ethereum. I don’t think that’s the case. In my opinion, the highest value transactions will continue to happen on Ethereum L1, and Rollup will expand the pie by providing ample space for users to trade on Ethereum. The relationship is symbiotic: L1 provides Rollups with low-cost data availability (I think it should be low-cost, and Ethereum’s security should be affordable to everyone), and in return, they make Ethereum L1 a natural intersection of all financial activity, where truly valuable transactions can take place.

Justin Drake: I personally think that ETH’s value accumulation comes down to flow and monetary premium. For flow, the important metric is total fees, not fees per transaction. As described in this talk, the end goal of Ethereum is to process 10 million transactions per second, which would provide billions of dollars in revenue per day even if the fees per transaction were less than 1 cent. For the monetary premium, the important metric is the percentage of ETH used as collateral currency, for example to cover defaults.

Anders Elowsson: In the long run, there is a direct connection between Ethereum promoting sustainable economic activity and the accumulation of value in ETH. If you design for sustainable economic activity, you are designing for the accumulation of value in ETH. If you want to accumulate value for ETH, you must specifically ensure that Ethereum promotes sustainable economic activity.

A short-term focus on "value accumulation" without considering the how and why may be detrimental in the long run. In some ways, although this is a sensitive topic and people may abuse me for my suggestions, I think the current roadmap is a roadmap that can accumulate value for ETH.

Anyway, I personally think that Ethereum’s success without ETH’s value accumulation is very surprising and perhaps a little disappointing, but I also think it’s a reason to buy ETH because I expect the market will eventually accept this view.

About the Ethereum Foundation

Q: What is the funding status of the Ethereum Foundation? How long can it last? What is the response plan when the funds run out?

Justin Drake: The financial report should be released soon. The Ethereum Foundation spends about $100 million per year (see this tweet from Aya). At current prices, the Ethereum Foundation’s main Ethereum wallet has about $650 million in assets. The Ethereum Foundation has enough fiat currency buffer funds to cover operating expenses for several years. Roughly estimated, the Ethereum Foundation has a 10-year buffer period. This period changes with the price of ETH.

Vitalik: The current rough budget strategy is to spend 15% of the remaining funds each year. That is, the Ethereum Foundation will exist forever, but its funds will decrease over time (as part of the ecosystem).

Q: Is the Ethereum Foundation research team equivalent to "core developers"?

Vitalik: There are many core developers outside the Ethereum Foundation. The most obvious example is the members of individual Ethereum client teams (such as Nethermind, Besu, Nimbus...). In addition, there are many independent researchers and contributors to specific topics (for example, some Optimism and Base people have made great contributions to the 4844 deployment).

Q: Is Ethereum development currently overstaffed, understaffed, or just right? If not, which areas need it most?

Vitalik: In my opinion, the entire P2P network aspect is not only understaffed, but also under-discussed.

Justin Drake: I think it is understaffed! If you are interested in helping Ethereum research, feel free to contact me via private message.

Other technical questions about Blob, Rollup, DeFi, etc.

Q: If Blob fails to reach the target average (3), should the target be lowered to ensure fee price discovery?

Dankrad Feist: Ethereum is currently creating a new Rollup market, the data availability market. Many alternative solutions want to take market share from Ethereum, including Celestia, Eigenlayer, Avail, etc. They can't compete on security, so they want to compete on price. Do we really want to artificially increase the price and have one of our largest assets (secure rollups) leave our chain?

At 3 blobs per block, Ethereum's protocol revenue will never break out. I think we should work hard to scale it as much as possible over the next few years. Then think about collecting fees from it.

I don't think collecting fees from blobs is the best value capture mechanism for Ethereum anyway. The data availability market is too fickle - while Ethereum provides the best security, it's not hard to get "close enough" performance, so it will never be a good way to capture value. Ethereum L1, as the natural financial intersection in the ecosystem, will have the highest value transactions, which I think provides the best value accumulation mechanism for Ethereum.

Justin Drake: Blobs will not fail to reach the target, we just need to be patient. Another factor worth considering is that Rollup projects (such as Base, Scroll, Taiko) have recently found ways to make better use of Blobs. These Rollup optimizations extend the time for blob price discovery.

Davide Crapis: It shouldn't. The mechanism is priced for congestion, so if there is no congestion, it is okay for the price to remain low. However, the current demand is far below the target value, which affects price discovery in the case of congestion. Therefore, price discovery is very important, and we should improve the efficiency of the mechanism. In the short term, raising the minimum fee (but still low!) or changing the update speed will help.

Q: In order to preserve the sorting fee, does Rollup have an incentive not to decentralize its sorter?

Vitalik: Actually, I don’t think it’s imperative for Rollup to decentralize the sorter. In my opinion, fee collection and sorter decentralization are orthogonal. If the sorter is centralized, then the Rollup sorts and then collects fees + MEV (but spends effort figuring out how to get MEV). If the sorter is decentralized, then the Rollup can get the revenue from auctioning the sorter position, and in equilibrium the revenue is equal to fees + MEV minus the cost of figuring out how to get MEV. The two cases seem symmetrical.

I think the main asymmetry is probably social. While the sorter is centralized, it’s easier to collect fees + MEV than to distribute it to token holders (or whatever distribution method is publicly agreed upon with the community). Decentralization requires “doing the right thing” economically. I hope that L2 will not continue to be centralized because of this, and I hope that the community (including organizations like L2beat) will take this into account and pay attention to this situation.

Q: What does the Ethereum Foundation think about DeFi? Do you think DeFi is the most valuable use case on Ethereum at the moment?

Dankrad Feist: Personally, I like DeFi, but it alone cannot solve all of Ethereum's problems.

Financial markets themselves do not create value. But DeFi does allow society to create more value through different functions, such as providing liquidity, insurance, etc. All of this is amazing. For example, one of the valuable things DeFi does on Ethereum is to provide us with decentralized stablecoins. Although the scalability of stablecoins is obviously severely limited, people are now more willing to use custodial alternatives. But I still think they are cool, especially for those who really need decentralized and censorship-free alternatives. It's nice to have decentralized stablecoins, but they can't scale to a billion users.

Besides that, I think what DeFi lacks the most right now is a set of "valuable" assets. In fact, I do believe that with the DeFi system ready to go, Ethereum has the potential to become the center of future financial activities, but there are still many things to build before that.

Q: If Bitcoin implements OP_Cat and develops a strong L2 ecosystem, what unique value can Ethereum provide?

Vitalik: Compared to Bitcoin, Ethereum has a larger Rollup DA space and more options for L2 security (Bitcoin has only 4 MB/600s = 6667 bytes per second, assuming that all on-chain data is used for DA); Compared to PoW, PoS has proven its decentralized capabilities and provides more options for 51% recovery; Ethereum has an efficient social layer, such as censorship panic, client centralization panic, pledge pool market share centralization panic, and many other things, which have been resolved through coordinated actions of the entire ecosystem; community, culture, values, etc.

Q: Why did Minimal Viable Issuance die? Given the low demand for blocks and blobs, shouldn't the Ethereum Foundation promote MVI (and MEV destruction) to let ETH ultrasound again?

Anders Elowsson: I would say that MVI is still alive and well. The current low fee burn puts the spotlight on issuance, which clearly shows that we do incur high costs under the current issuance system, and these costs can be reduced to make Ethereum better. However, MVI is valuable in its own right and is not a tool specifically designed to make Ethereum "ultrasound".

MVI achieves two key points, including that reducing issuance reduces the costs that Ethereum users have to bear (hardware, risk and opportunity costs, liquidity losses, taxes, etc.), thereby improving overall welfare (see also this blog post). In addition, if everyone stakes, some LST may become "too big to fail" and the social layer may struggle to maintain the expected consensus process. The surge in LST will also hinder ETH's status as a trustless currency, making Ethereum unattractive due to the risk of future monopoly pressure.

For these reasons, MVI is desirable in both 2% deflation and 2% inflation scenarios.

Q: Should ETH be net deflationary in the long term? Before EIP 4844, users paid high fees and ETH was deflationary. After EIP 4844, users paid low fees and ETH was inflationary. How can we achieve both deflation and low fees in ETH?

Anders Elowsson: Both requirements can be achieved through scale, which is conducive to sustainable economic activity. To achieve scale, millions of users pay low transaction fees and have transactions secured by Ethereum. This is a sustainable way to increase total fees. Barnabé's answer three years ago explained this quite well from a user perspective.

Speaking of the situation before and after EIP 4844, I would like to add that this cannot be simply attributed to the change itself. L2s were developed on Ethereum and paid high fees to "settle down" on Ethereum because Ethereum had a plan to reduce fees. Without this plan, these L2s may never be developed. Similarly, if the plan is not implemented, these L2s may eventually leave and go somewhere else. The gas price at a certain point in time reflects everything that has come before and everything that will come. This shows that in the long run, abandoning expansion will not necessarily make Ethereum net deflationary, because transaction demand is partially driven by future expansion commitments.

Another point worth mentioning is that perpetual net deflation requires more than just a reduction in issuance or an increase in burning fees. Under a reward curve that adapts to the amount of staked rather than the staked ratio, the long-term staking equilibrium will eventually be affected by the circulating supply equilibrium as well, because the circulating supply will change to balance supply, demand, and protocol revenue. To achieve perpetual deflation, the issuance policy must be changed by changing the reward curve and normalizing the circulating supply when swapping. Once the circulating supply begins to be tracked at the consensus layer, perpetual deflation can be achieved.