Virtual Blockchains in The Modular Era

The pioneers of new technology must raise a lot of capital to create foundational infrastructure, which can lead to over-investment and speculative bubbles. When these bubbles burst, weak firms fail, and market power consolidates around industry leaders and their paradigms. Through this consolidation process, we can identify the common elements across applications and isolate them into standard, modular components that can be open-sourced or sold as individual services. These abstractions make it easier to build more complex applications and enable a shift from capex-dominant to opex-driven cost structures that allow new products to launch faster and with lower startup costs. This pattern is now unfolding in web3 as new “modular” technologies, such as rollups, accelerate development and unlock an era of lean startup innovation. 

Capex to Opex

Technology infrastructure becomes more powerful and easier to use as it becomes more standardized and broadly available. But until that happens, early entrepreneurs must invest heavily into building their own infrastructure before they can distribute its applications, like Edison inventing the grid to sell the lightbulb or early internet startups deploying data centers to run web pages. As the market matures, open standards and on-demand infrastructure services emerge, enabling more efficient business models for companies who adopt them because they don’t need to spend as much time and money to bring their products to market. 

After the dot-com crash in 2000, for example, the internet industry transitioned from buying servers and setting up data centers (capex) to renting them from the cloud (opex). Many open-source frameworks like the LAMP stackRuby on Rails, Django, and NodeJS emerged to simplify web development, while industry leaders like Microsoft, Amazon, and Google leveraged their scale to establish new standards and low-cost infrastructure services. This co-evolved with the API boom that started in the late 2000s, which further simplified the complexities of the internet by offering specialized backend functionality under pay-as-you-go business models. Within a decade of the crash, these abstractions allowed small teams to build and scale new applications quickly and cheaply, accelerating innovation and fueling the golden age of startups that defined the era. 

Web2 infrastructure became so abstract that modern web applications don’t even run directly on physical servers but in simulations of servers: virtual machines, often wrapped in containers that can be easily moved or replicated across many environments with minimal reconfiguration. This technology helped scale web2 by allowing a single, powerful server to run many applications simultaneously and makes it easy to add or subtract compute resources to applications as needed to match demand and control costs.

The concept of virtualization illustrates how abstract infrastructure can become, but I highlight it here because web3 infrastructure is following a similar path with the invention of rollups, which similarly helps blockchains scale by allowing them to power multiple “virtual blockchains” on top.

Layers of Abstraction

Early blockchain startups had to build all their infrastructure – including custom blockchain protocols, frontends, wallets, SDKs, APIs, and more – just to get started. Smart contract networks like Ethereum reduced the need to build proprietary blockchains for many applications, but they imposed significant constraints on costs, programming conventions, and scalability, limiting the universe of possible applications. The more ambitious ideas required a level of flexibility and throughput typically unavailable on public chains, so many of the most exciting applications were impossible to scale. 

Platforms like Cosmos and Polkadot later provided tools to create custom blockchains with shared security and interoperability features, making it easier to launch secure chains. However, they still demand substantial resources and expertise to use and thus remain out of reach for most developers. But just as more layers of abstraction simplified the cloud, emerging Layer 2 (L2) standards like rollups allow developers to deploy blockchain environments quickly and cheaply. 

Rollups execute transactions and smart contracts off-chain and bundle the results of multiple operations into periodic, cryptographically verifiable transactions on a primary blockchain, thereby inheriting the security of the underlying network. It’s akin to how credit card networks process many payments and settle them in weekly batched wire transfers to merchants. With this technology, a single blockchain can secure many higher-performance virtual blockchains simultaneously, drastically increasing throughput capacity while minimizing transaction fees.

Importantly, rollups are not blockchains, at least not any more than virtual machines are actual machines. Rollups are virtual blockchains, simulated environments where smart contracts run as if on a real chain, oblivious to the abstraction. A rollup can be as centralized as needed for performance, control, or compliance as long as the operator frequently settles the output on a trusted blockchain and is disincentivized from corrupting the data. But it can also be decentralized by using “shared sequencer” technologies. 

Besides scaling, unbundling the “execution” layer from the “data availability and settlement/consensus layers gives developers the flexibility they need while leveraging the security guarantees of the main chain. You could, for example, deploy an application as a rollup that uses Python as a programming language but settles to Ethereum if you dislike Solidity yet want to leverage Ethereum’s security or ecosystem. Open-source frameworks like the OP Stack, the ZK Stack, Polygon’s CDK, Arbitrum’s Orbit, or Rollkit, to name a few, already make it easy for developers to deploy custom rollups with different levels of trust, while decentralized sequencer projects like Espresso and Astria give the option to decentralize the execution layer if necessary. Meanwhile, a growing number of low-code “Rollup as a Service” (RaaS) products, like Dymension, Conduit, Caldera, and Gelato, allow anyone to launch custom virtual blockchains in minutes. 

The broader “modular movement” further empowers developers with co-evolving standards and services covering other areas of the stack, further reducing the costs of building and scaling blockchain applications. Ethereum’s EVM dominates as an “operating system” for smart contracts, while Solana’s SVM is rising quickly as a high-performance alternative (both can be used in independent rollups). Protocols like POKT standardize the RPC/API layer across networks, while platforms such as Syndicate abstract all the complexities of building sophisticated blockchain applications behind simple APIs any developer can use. Frameworks like Polywrap abstract multiple protocols into a single front-end SDK; bridges like Across enable liquidity to move between networks, and wallet standards like SAFE or Squads plus “Wallet as a Service” (WaaS) companies like Magic make creating custom wallet experiences for users on any chain easy. There’s even a new crop of L1s like Celestia that are purpose-built for virtual blockchain environments.  

Millions of Virtual Blockchains

Many believe high-throughput networks like Solana can reach “web-scale” without L2s, but people vastly underestimate what web-scale means because most of the activity on the internet happens in the background. Every click you make triggers hundreds of hidden HTTP requests; just loading triggers over 300 background requests to different APIs and service providers in under 2 seconds, and that’s for one user’s single action. Achieving web-scale means reaching millions of transactions per second, per application, and even that’s not enough if internet demand multiplies by a million as AI proliferates. Virtualization is necessary to reach that level of scale, but we also need ultra-high-performance L1s underneath to achieve it. Besides blockchains optimized for data availability throughput, like Celestia, high-performance networks like Solana and Monad are potentially interesting playgrounds for rollups.

That said, scalability isn’t the only reason virtualization is important. Virtual blockchains are a powerful standard for the web3 generation of online services. The first wave of rollups mainly consists of “faster Ethereum” services. However, the flexibility provided by a modular architecture makes virtual blockchains particularly useful for creating application-specific environments or networks tailored for specific ecosystems, industries, or geographies. You can also create “virtual private blockchains” for use cases with strict access control or compliance requirements. Bigger yet is the idea that virtual blockchains could become the default backend infrastructure for all online applications, as Blockchains and Smart Contract Interfaces replace web2’s “Cloud and API” paradigm (the subject of my next essay). 

The current strategy for web3 startups is to first launch on high-performance, low-cost networks (like Ethereum L2s or Solana), with eyes on migrating to custom, application-specific environments if they need more scale. Even existing protocols that had already built their own chains, like Celo or POKT, are transitioning to an L2 architecture to streamline their infrastructure costs, echoing the time when internet companies with data centers had to adopt the cloud as a matter of business. If you don’t embrace the new, you’re vulnerable to competitors who do.

What happens as the marginal cost of deploying a secure virtual blockchain environment approaches zero? For one, it perpetuates consolidation at the base layer. L1s are capital-intensive by design because that’s how they provide security guarantees; the high costs of successful L1s make them immensely valuable and hard to compete with. Market leaders will only strengthen their position as more firms opt to use virtual blockchains on top of established networks instead of creating new L1s. The market will likely consolidate around 3-5 major blockchains backed by trillions in capital that serve as global security, liquidity, and interoperability infrastructure for millions of virtual blockchain applications, protocols, and digital assets. New, highly valuable L1s will continue to emerge, but innovation and outsized returns will mostly move up to the application layer

We will go deeper into these ideas in subsequent articles, but the most important point I want to highlight from a business perspective is that modularity represents web3’s shift from capex to opex, and consequently, we can expect a rapid expansion of next-generation blockchain applications. Opex means costs scale with growth instead of being front-loaded by large capital raises before launch. It means entrepreneurs can iterate faster, applications can scale cheaply, and investors can finance ventures with less risk. Like the web after the dot-com crash, these are prime conditions for a golden age of startup innovation in crypto. 

Artificial Intelligence Belongs Onchain

As the cost of producing artificial intelligence models decreases, the population of AI agents will grow exponentially. Agents will soon outnumber humans online, creating, consuming, and exchanging multitudes more information than humans ever could. But if we get, say, a million-fold increase in digital activity, and 99% of that growth comes from machines, it will be hard to cope with this transformation without adopting onchain infrastructure and business models that both empower agents to reach their full potential and allows us to identify, control and audit their actions. 

Today, companies like OpenAI take on the massive costs of producing models and then sell access via their proprietary interfaces and APIs, and they’re mostly limited to consuming and creating content. But to fully harness the potential of AI, we need a vast population of specialized agents that can talk to and transact with each other. They must be free to roam the internet and be able to have and spend money to execute tasks on our — or their own — behalf. We also need ways to identify, control, and audit their actions. 

The problem is that we can’t coerce agents to follow our laws, preventing effective regulation. They can’t use the traditional financial system, which depends on a jurisdictional identity model, limiting their ability to transact. They consume a ton of online information — imposing all the traffic costs on service providers — but don’t generate the corresponding revenues by subscribing or clicking on ads. To solve these problems, we need a natively digital legal and financial system combined with new business models to take full advantage of the opportunities presented by this new technology. 

The solution requires (1) sovereign digital infrastructure with a new software paradigm that guarantees trusted code execution with an immutable audit trail; (2) an independent, digital financial system that treats humans and machines alike; and (3) a cryptographic identity model coupled with decentralized communication and reputation protocols. This is only possible using blockchain protocols and smart contract applications. 

Blockchain protocols provide all sorts of decentralized digital services that are accessible via smart contracts and payable with digital assets. For example, smart contract networks like Ethereum and Solana enable secure, reliable execution of open-source software backed by an auditable trail of blockchain transactions, and networks like Filecoin and Arweave provide cheap and scalable onchain data storage services. As it becomes easier and cheaper to build new protocols on top of established platforms, the range of services offered via decentralized networks is expanding.

We can use these platforms to train, deploy, and operate agents in a decentralized manner, but more importantly, the ability to consume them via smart contracts facilitates agent interactions. It’s much harder for AI agents to consume a typical Web2 REST API than to read a smart contract and pay for its service with tokens – no accounts or credit cards required. 

Wallet-enabled agents can use any smart contract service or platform, from infrastructure services to DeFi protocols to social networks, which opens a whole universe of new capabilities and business models. An agent could pay for its own resources as needed – whether computational or data – which is a big idea. It could trade tokens on decentralized exchanges to access different services or leverage DeFi protocols to optimize its financial operations by borrowing opportunistically or earning yield from its assets. It could vote in DAOs or even charge tokens for its functionality and trade information for money with other specialized agents. The result is a vast, complex economy of specialized AI agents talking to each other over decentralized messaging protocols and trading information onchain while covering the necessary costs. It’s impossible to do this in the traditional financial system. 

Consider the consequence of this idea: if agents act onchain – even if they think offchain – we end up with a public, immutable and cryptographically-signed record of their activity over time. Blockchains will ensure the safe deployment of AI at scale, enabling us to do things like audit agents’ internet actions, distinguish machine-made from human-made content, and build identity and reputation systems for machines based on their onchain activity. It will help us (and them) differentiate between good and bad actors or tell which agent performs better than another at a particular task. Agents will then be able to decide who to rely on based on their onchain history, which they can easily access thanks to the open-source nature of blockchain data.

Many pieces are coming together to enable this vision. Blockchain infrastructure is rapidly becoming fast and cheap thanks to new consensus mechanisms and scaling solutions.  Smart contract wallets and “wallet as a service” (WaaS) providers will enable agents to transact; meanwhile, emerging account abstraction techniques can allow human-agent interactions where we can authorize agents to spend from our wallets. We can build registries and reputation systems (including blocklists or slashing mechanisms, for example) using agents’ public keys as identifiers. We can even play with DAO-owned agents and experiment with new business models; perhaps some agents will live on their own Layer 2 network, owned and managed by a decentralized community of operators. 

These are the kinds of ideas that initially seem far-fetched but turn out to be very obvious in hindsight. Of course, smart contracts will mediate commercial activity between agents just as legal contracts govern that of humans. Of course, agents will use the internet’s financial system powered by digital assets instead of banks and credit cards. Of course, agents will identify, communicate, and transact with each other using cryptographic identities via decentralized protocols. I don’t see another way around it: artificial intelligence belongs onchain. 

Ethereum and Solana

Ethereum and Solana are like Android and iOS. Android values modularity: it runs on many different types of devices made by hundreds of manufacturers worldwide; Google only makes 1-2% of them. This approach made it the world’s most popular mobile operating system, with an estimated 60-75% market share. Android’s flexibility has been a boon for hardware companies making anything from smartphones to televisions, as they can bring new products to market without investing billions into building bespoke operating systems. However, such diversity also makes it more difficult to develop apps that seamlessly work across many devices with different specs, screen sizes, and the various versions of Android these devices run.

By contrast, because Apple makes all iOS devices, it can provide users and developers with a more integrated and consistent experience. The time saved by not having to optimize across different devices can go into delivering better apps that users are willing to pay a premium for, and it’s not uncommon for companies to launch on iOS first as a result. So, while Apple has just about a third of the market in terms of distribution, it does a much better job at capturing the value of its ecosystem with a whopping ~60% of all mobile spend, plus all the hardware revenue. 

It’s not an original analogy, but it’s a useful one. Ethereum is similar to Android in that it’s quickly becoming more of a platform for third-party networks than the place where most end-users and developers operate. The fast-growing ecosystem of layer-2 (L2) networks using Ethereum mainnet for security or leveraging the Ethereum Virtual Machine (EVM) as an operating system already processes ~5-6x more transactions than Ethereum mainnet, and it’s easy to see how that number can quickly expand to 100x and beyond. These third-party networks are like the different Android OEMs: many offer directly competing services with slight variations, while others focus on specific markets or use cases. With a growing number of tools making it easier to launch L2s, we can easily imagine Ethereum’s “sphere of influence” growing to encompass hundreds of networks processing billions more transactions than Ethereum mainnet. 

The EVM is poised to remain the most popular blockchain operating system, running on thousands of different networks and rollups across markets, sectors, and use cases. Ethereum will significantly benefit from this expanding ecosystem even as mainnet captures just a portion of its total value. However, this diversity brings many of the same challenges Google has with Android. For example, different EVM networks can run slightly different operating system versions, so Ethereum smart contracts aren’t guaranteed to run seamlessly on all of them by default, and developers have to spend extra time tweaking, testing, and maintaining for different environments. The Ethereum user experience can also become too fragmented for the average user: apps on one network may not be available on another, wallets don’t support all networks simultaneously, and switching or bridging across them can be confusing or even dangerous for the average user. These UX issues will get smoother over time, just as Android got smoother and safer as it developed. Still, the burden remains on developers to invest more time accounting for these issues than they would have otherwise.

On the other hand, Solana is similar to iOS in how it values tightly integrated components in the name of throughput and performance. There’s much more to it – different consensus mechanisms and design principles – but ultimately, Solana as a single, unified network is much faster and cheaper than Ethereum and many other EVM networks. Developers can focus on delivering better apps for a single, high-performance platform without worrying about transaction speed, gas cost optimization, or deploying across multiple networks, and users don’t have to worry as much about slow transactions, switching or bridging between networks, inconsistent wallet support, and other usability issues that plague Ethereum. 

Of course, this is a surface-level analogy, and much detail is lost in the comparison. But it’s helpful to understand that it’s a game of nuanced choices, not winners and losers. Which is absolutely, objectively “better” doesn’t matter; that debate will never end. Some people will value a modular philosophy’s flexibility; others will value a more integrated platform’s speed and simplicity. What matters is that options exist, and you can choose what works best for you. The success of our industry requires a competitive set of platforms with different tradeoffs to maximize developer and user choice; today, Ethereum and Solana embody this spirit as two of the most promising smart contract networks in the market.

Understanding where we are in the financial and technology cycle is essential for investors. Bear markets are the best times to survey the market and pick new winners during a significant consolidation. Maxis would have you support only Ethereum, Bitcoin, or whatever network they favor and discard everything else, but that’s more emotion than logic. Indeed, there was a time when BTC was the only legitimate digital asset, and there was a time when Ethereum, like IBM at its peak, was the only real game in town despite emerging alternatives. But it’s a mistake to assume that the status quo at any point in time will remain the same forever.