Really interesting debate among prominent figures in the Crypto space on a topic that seems extremely relevant to understanding the fundamentals of Layer 1s

Let's start with one of the main questions that arise about the demand for deccentralized blockspace.

“I find it very convincing that long-term, (decentralized) blockspace demand is highly elastic (and very unlikely to be inelastic). Are there any good economic arguments for or against this, other than looking at some use cases that become possible at certain prices?” Dankrad Feist

It's important to clarify that we're talking about decentralized blockspace, and that definition is extremely complex since we can't talk about decentralization as something binary, but rather as a spectrum. Perhaps the analogy that seems easiest for me to understand is with the demand for cars, specifically the demand for luxury cars.

There is an alternative debate about whether blockspace is a commodity. However, I think there is a high consensus, or at least the market values it in such a way, that this resource is not really homogeneous. That explains why the market is willing to pay much more for space on a fully decentralized network than on another that is on the opposite side of that spectrum. We could perhaps argue that decentralization is not the only property we should consider in blockspace, and that’s true since another important factor is the number of active users on that network, for example. Therefore, I would even say that it's not easy to value each type of blockspace, and that valuation changes as these networks evolve.

But from a demand perspective for this application deployment architecture, there seems to be a clear trend: the demand for blockspace is exponential. This is a really solid trend explained by many factors, including, of course, the ability to have digital assets, but also operational costs and distribution to the users of these platforms. Therefore, I would say that, from a general point of view, the demand for blockspace is highly elastic based on the current data we have, placing us in a scenario similar to the demand for bandwidth on the Internet.

Continuing with the car analogy, it's possible that the demand for cars is very elastic, but is the demand for luxury cars equally elastic?

The demand for less demanding blockspace would place us in a system very similar to traditional ones, where the cost would probably be close to that of cloud computing services, with perhaps some additional cost for the extra infrastructure needed to provide the service. This type of demand might be met by higher layers, such as L3 and L4, which would be much closer to traditional cloud services and where the value proposition doesn't seem very different. However, as these architectures begin to offer many services, even if centralized, they also start to add value through the ability to use open-source software.

“Arguments for highly inelastic:

• Financial services that dominate decentralized blockspace are often essential & inelastic

• Decentralized blockspace is very unsubstitutable

• Network effects & economies-of-scale (esp. liquidity) are profound

• Highly specified” Polynya

Financial services that dominate decentralized blockspace are often essential and inelastic. I do not agree with this argument and believe it is important to differentiate purely transactional services from capital management services or those associated with some risk. Purely transactional services, in my opinion, would likely develop into more transaction-intensive services if there were a greater supply and, therefore, lower prices. For example, sending money per second would be one of those services that would emerge with marginal transaction costs.

The arguments about the scarcity of decentralized blockspace and its network effects are, network effects, to me, more like characteristics of this high-value decentralized space. It's true that before the explosion of higher layers, it was the only possible space, but now with this ability to scale, we have options of somewhat lower quality at a much lower price. This actually shifts certain demand to much more competitive alternatives that didn't really require such a expensive decentralized space.

In other markets, the response has been that the large volume really lies in offering mass services at marginal prices. While there is a demand for luxury services, the size of that market is smaller. However, it is still a significant market

This article by Vitalik is really interesting on this topic explaining the main applications that demand this decentralized blockspace.

What in the Ethereum application ecosystem excites me

Hence my change in perspective: my excitement about Ethereum is now no longer based in the potential for undiscovered unknowns, but rather in a few specific categories of applications that are proving themselves already, and are only getting stronger. What are these applications, and which applications am I no longer optimistic about?

https://vitalik.eth.limo

To discuss this topic, it's interesting to analyze the demand data currently provided by Ethereum. One of the first conclusions we can draw if we analyze it historically is that this demand has changed significantly

Let's delve deeper into the claims made in Polynya's article.

Money is one of the main applications highlighted in the article. There is no doubt that if we had to talk about a killer application of public blockchains, stablecoins would probably be the most important. Indeed, in the rankings, we see USDT as one of the most used contracts within the Ethereum network.

However, I think we could consider that this type of demand fundamentally comes from a more general need, which is the creation and transfer of digital assets. The movements of ETH could be considered in the same category. In this category, we probably have a long tail of transactions, which would include all those related to ERC-20 tokens, both in creating these assets and transferring them. The main value proposition of Ethereum compared to other networks is the 'legal certainty' it offers as a platform for digital property rights.

One of these categories, and probably the most successful, would be stablecoins. The demand for transactions with these assets is clearly impacted by the transaction costs. In this sense, the demand for low-value transactions is completely affected by the transaction cost. This demand has shifted to higher layers for low-value economic demands, limiting this network to high-value payments of both stablecoins and other types of assets. However, the demand for the latter depends more on adoption than on transaction costs. Nevertheless, if we envision Ethereum becoming the main platform for asset registration, there is much room for this demand to grow. If the direct capacity of L1 could be expanded, it would be normal to see more transactions on Ethereum, as it would offer greater guarantees at the same price.

DeFi, however, offers a greater need for space and transactionality per service. In this regard, Uniswap currently leads the demand for decentralized blockchain. The exchange of assets is probably one of the network's main activities. Aggregators also occupy a significant portion of the demand, optimizing these exchanges. However, we once again exclude the demand for low-value exchanges for the same reason mentioned earlier. In this case, we should also add the limitation of financial applications that are very transaction-intensive, such as high-frequency trading. However, market liquidity generation operations, like liquidity pools, which are less demanding in terms of transactionality, are viable. What does seem reasonable is the idea that Ethereum could become, at the very least, the registry where all transactions carried out on other layers or sidechains are settled.

Identity This is a category that has not yet emerged as one of the actors that may demand more blockspace. The creation of Ethereum accounts does not require much space, and the continuous demand is more for reading. In this category, we might find projects like ENS or POAP, but many of these projects are not compatible with high transaction costs. However, it's really important to try to establish identity within the network because this directly impacts the number of user assets and, therefore, the network effects. We are also talking about an environment where one of the core values is universality. This means that we shouldn't always be able to require identity.

DAOs This is one of the main sources of demand for decentralized blockspace. Not so much because the activity is very frequent, but because it indirectly fosters an environment for creating economic activity. Considering that in the future we may use these platforms as the heart of governance for technological projects, the demand could be really high. A large part of the digital assets defined are shares in the governance of these projects. Another major success is creating a jurisdiction that has generated enough trust to invest globally. There is much more friction for an American fund to invest in a jurisdiction outside the US than to invest in tokens on Ethereum, as they have a certain capacity to access the assets of that company by being digital and subject to voting.

Autonomous agents are probably one of the most powerful sources of demand we could have. In this sense, aggregators would be a primary example of this demand since they perform many transactions on behalf of the user who initiates them. AI could be one of the major sources of demand in this environment. However, the transaction cost will determine the level of economic activity that is viable in this context.

We could argue that purely human transactionality has a limit in terms of the number of transactions, but if these are performed by agents or programs, there is really no limit beyond the economic viability of the activity. In traditional markets, most trading is done algorithmically. In a more sophisticated financial environment, it makes sense for our orders or intentions to deploy hundreds of microtransactions. The trend should be to bring all this transactionality on-chain, but a layer like L1 cannot offer a marginal cost for all these transactions.

We must not forget that while connectivity is one of the main competitive advantages of this ecosystem, bringing financial services to marginal transaction costs is also crucial.

NFTs, Memecoins and Speculative Phenomena

Ethereum has a complex relationship with these phenomena. Its deep social and ethical beliefs, which in my opinion underlie the community, lead it to inherently reject these types of activities. Even in the economic design of Ethereum and its issuance, there is a noticeable internal struggle between the more technological side, which doesn't prioritize the asset's value, and the more economically-focused research team that emphasizes the importance of value capture. All of this takes place in a context where the primary use case of the network is financial, and where the value capture of the token has, for the first time, successfully combined open source with financial resources. The most demanding contract in decentralized blockspace is Uniswap, which is for trading. Historically, NFTs have probably been one of the activities that most strained the network, sometimes even causing it to become congested. NFTs have lost their speculative moment, giving way to memecoins, and this activity has generally shifted from being the main one on Ethereum to moving to networks like Solana and recently to starting to develop on L2 solutions such as Base. Perhaps if there is still some demand for this activity on Ethereum, it might be for high-value NFTs. However, memecoins, although they have a huge market capitalization in aggregate, are typically small transactions, which pushes them away from L1. The theory is that this activity should take place on L2, allowing L1 to capture the settlement of these operations. Nevertheless, with the introduction of Blobs, settlement on L1 has become so cheap that it seems unlikely that this activity will regain a similar role as it had for a long time. This is all in an ecosystem that does not align with this type of activity for very understandable reasons.

In this space, we have a competitor that we perhaps didn't expect: Bitcoin. It hasn't fully developed yet, but it is certainly poised to compete for high-value transactionality, having natively defined both tokens and NFTs

MEV

There are many voices arguing that the business model of L1 networks may not be based on transactions but rather on activities like MEV (Miner Extractable Value). Indeed, MEV has become one of the largest gas consumers, and it seems that this type of value extraction should ideally be capitalized on by the networks themselves. However, in most cases, it effectively acts as a sort of commission on economic activity, as many of these activities ultimately impact their indirect costs. This is especially true in trading, where it occurs most frequently. In this sense, MEV can be seen as a hidden fee for users, which, in a competitive environment, should also be optimized. However, in some cases where this activity doesn't merely extract value but can even be positive, it could add a significant new source of revenue. In such cases, MEV could function like an agent that not only profits but also makes economic activity more efficient. Nonetheless, this is not the most common scenario.

Summing up:

My conclusion is that the demand for decentralized blockspace is exponential, and in the short to medium term, it is particularly driven by adoption. On-chain offers an environment with many competitive advantages, leading us to believe that it will be the future of application deployment. Decentralized Blockspace cannot be considered a commodity, as not all execution spaces offer the same capabilities. Decentralization is certainly one of the main value propositions. However, at this stage, decentralized blockspace is already beginning to offer a wide range of options with different levels of quality and at different prices. It is in this space that different types of activities are beginning to find their place. It is a highly competitive environment, where we will increasingly see the offer of quality space at lower prices, within the limitations imposed by the costs associated with different architectures.

Ethereum positions itself as the most decentralized and neutral blockspace, currently leading in terms of resource expenditure, though not necessarily in terms of quantitative demand, which we might find on other networks. The Ethereum ecosystem as a whole is undoubtedly the leader in aggregate, with ecosystems like Solana and Bitcoin as other major alternatives. Therefore, we might say that certain demand for Ethereum is inelastic, as long as it needs to be on the most secure network with the most capital or network effect. However, the rest of the demand will find its place based on transaction price and other alternative characteristics.

It is also challenging to predict the impact these demand changes will have on network effects, as the fragmentation of the Ethereum ecosystem is a handicap for certain activities. However, we could argue that this demand is similar to that of the Internet, as its exponential trend does not seem likely to reverse, especially at a time when cost optimization invites activities that were previously unviable. In any case, these activities will always begin on low-cost networks. The key will be understanding how these activities can settle on L1s like Ethereum or if, in the end, the networks that dominate demand aggregation will be able to abstract the settlement layer in the future.