The Agent Era Reshapes the Logic of Blockchain Value Distribution

Who Profits from Agents?

Original author: Jonah Burian

Peggy

Editor’s Note: If agents truly become the next billion users of blockchain, the more important question may not be “How much transaction volume will they generate?” but rather, who will profit if this world comes to pass?

In the past, whether the "fat protocol" or "fat application" theory, it was assumed that on-chain users were human. Humans care about whether the interface is user-friendly, whether the brand is trustworthy, and whether the path is convenient; thus, the application layer could capture value by controlling user access and transaction flows. But agents are different. They directly call APIs, have no brand loyalty, and can switch between different protocols, aggregators, and exchanges at low cost.

This means that agents may rewrite the value distribution logic of Web3. The application layer could shift toward a "headless" model, exposing wallets, aggregators, and deposit/withdrawal capabilities as APIs for agents; the protocol layer might also regain opportunities as agents bypass intermediate layers; but in more radical scenarios, agents could push the entire on-chain stack into price competition, compressing profit margins for applications, aggregators, and infrastructure down to near marginal cost.

What truly matters is that agents don’t just increase the frequency of existing on-chain transactions—they may create entirely new activities that weren’t viable before: continuous portfolio rebalancing, machine-to-machine payments, and new types of markets that only make sense with automated, high-speed execution.

Therefore, the core question of the Agent era is not simply whether value will flow to protocols or applications, but rather who can convince Agents to return here despite having infinite alternative options. The answer may no longer be UX and branding, but rather liquidity, latency, settlement finality, or some new business model not yet named today.

The following is the original text:

Many people envision agents becoming the next billion users on blockchain. But few ask the follow-up question: If this world truly comes to pass, who will make money?

Historically, all theories about value capture in the crypto industry assumed users were human. The "Fat Protocol" theory posits that the protocol layer is best suited to monetize users. In contrast, our "Fat Application" theory, presented in our works "How to Capture Value" and "The Great Repricing," argues that the application layer performs better.

But the Agent changed who the "user" is, rendering existing value capture theories unreliable.

In 2016, @jmonegro wrote "Fat Protocols." Since then, nearly a decade later, this article has become the dominant value capture theory in the crypto industry.

The core idea is this: In the internet era, value primarily flowed to the application layer—such as @Google and @Facebook—while the underlying protocols, like TCP/IP and HTTP, captured almost no value. But the crypto industry reverses this dynamic. Since blockchain data is open and shared, applications become commoditized; meanwhile, the protocol tokens required to use the network capture speculative value as usage grows. Every successful application drives increased demand for the token. Ultimately, the protocol layer will compound faster than any application built on top of it.

For a long time, this reasoning appeared correct: the market capitalizations of Bitcoin and Ethereum exceeded those of any companies built on top of them. This model held because, at the time, the protocol layer was scarce, expensive, and difficult to replace. In 2017, both Bitcoin and Ethereum were indeed scarce, as there were not yet a dozen general-purpose L1 blockchains competing for the same workloads. Block space was so limited that holding the underlying asset was akin to owning a stake in every application that required the network.

But now, trusted alternatives exist at every layer of the infrastructure stack: multiple high-throughput L1s, dozens of L2s, and modular settlement and data availability layers competing on price. Block space has shifted from scarce to abundant. As cross-chain bridges and aggregators make the underlying chains nearly invisible to users, switching costs are rapidly declining. Infrastructure has become interchangeable, and interchangeable things ultimately compete only on price. Consequently, pricing power at the protocol layer has vanished along with scarcity.

By 2026, the entities capturing significant economic value will no longer be protocols, but applications such as @phantom, @coinbase, @Polymarket, and @Pumpfun.

In my view, the reason is that the most valuable asset in the crypto industry is user relationships. If you control the user interface and transaction flows, you control distribution; and as long as users interact with any on-chain product, you can monetize it in nearly every way—exchange, lending, staking, minting, deposit and withdrawal channels, and more. This is likely also why investment firms are so obsessed with neobanks.

The app will also push infrastructure into pure price competition, compressing infrastructure profit margins to near marginal cost. I documented this strategy in "How to Capture Value." The same dynamic is now occurring in the stablecoin space, which I also discussed in another article.

The price is reflecting this theory. Spencer and I call this shift The Great Revaluation: during this cycle, value flows to the layer that owns user relationships.

The "fat apps" theory assumes users are human, and humans value user experience, branding, and convenience. But agents do not prioritize these; they directly call APIs, have no brand loyalty, and can switch trading venues at zero cost.

When users become software, owning user relationships is no longer as defensible. The entire "fat application" theory, which relies on a frontend moat,随之贬值.

In the Agent era, who will capture value?

The application is becoming headless.

One possible future is that application-layer winners will continue to remain winners, just without the UI.

Wallets and aggregators have already built the most challenging parts: the ability to integrate with a wide range of protocols, routing logic, identity, and fiat on/off-ramp infrastructure. The natural next step is to expose this capability as an API for Agents, enabling them to perform routing just as human users today do through @phantom or @JupiterExchange.

In this world, the "fat applications" theory still holds, but without the frontend. Companies that triumphed in the era of human users will re-platform themselves as headless infrastructure. We are already seeing traditional SaaS companies like Salesforce moving in this direction.

Protocol resurgence

Another possibility is that the Agent will completely bypass the intermediate layer.

If integration is simple enough—such as with clear API documentation, standardized RPC, and predictable execution semantics—there is little reason for an agent to pay an aggregator to perform tasks it could easily accomplish on its own.

Aggregators gained advantages in the human user era through user experience and complex routing capabilities. However, agents do not require user experience, and routing itself is an engineering problem that can be solved—and agents are increasingly adept at handling such tasks.

If the future unfolds this way, the "Fat Protocol" theory will experience a second life.

Pricing power across the entire stack will collapse.

Another possibility is that the Agent will impose commoditization pressure across the entire stack.

They are sufficiently rational. They always choose the cheapest trading venue, with no loyalty and no friction. Apps will lose the UX premium they previously charged human users. Aggregators and infrastructure will also lose pricing power, as there will be no human user inertia to shield them from price competition.

In this scenario, it is difficult for any layer in the stack to capture significant value. The entire supply chain will be compressed to near marginal cost, and economic surplus will flow to the party owning the Agent, or to the end users represented by the Agent. Cryptocurrency will become a utility, and utilities are typically not easy places to make money.

The agent will create new activities that were previously not feasible.

The simple version of this view is: Agents will do what humans already do, just at a much higher throughput; even if profit margins shrink, a significant increase in trading volume will still make the overall pie larger.

But I think there’s an even more interesting version: Agents will make a class of activities previously impossible now feasible—such as continuously rebalancing portfolios with execution costs below one cent; machine-to-machine commercial transactions between agents; and markets that only make sense when pricing and trading speeds outpace what humans can realistically keep up with.

These activities will not appear in our framework for observing on-chain activity today, as we assume that human participants are always involved in on-chain activity.

If this is the true change brought by Agent, then the question is no longer how to divide the existing pie, but how much new economic activity will be brought on-chain, and which layers are best suited to serve these new activities.

In each cycle, we try to guess where value will flow, often assuming that existing business models will naturally extend into the future. But this assumption typically misses business models that haven’t yet emerged.

When the internet was first built, no one predicted the emergence of the attention economy. The business model that now seems obvious—dividing users’ attention into fragments, auctioning them off to advertisers, and allowing a single company to take a substantial share of global advertising spending—was entirely unfamiliar at the time. It only appears inevitable in hindsight.

AI appears to be one of the biggest technological disruptions in decades. In an agent-dominated world, a portion of value capture will likely flow to business models that haven’t yet been seriously considered today. The participants who ultimately capture value may not be the ones currently receiving market attention.

The most likely outcome is not one paradigm completely replacing another. Humans and agents will coexist as users of the crypto industry for a long time, and the value capture maps for these two types of users are different.

As long as humans remain directly interacting with blockchains, the "fat applications" theory still holds: consumers willing to pay for user experience, brand, and convenience will continue to pay a premium to applications that have established user relationships. Meanwhile, at the layer where agents conduct transactions, a different set of principles will govern—specifically which one depends on how the above scenarios ultimately evolve.

In my view, the most important question builders should repeatedly consider is: What would make an agent return to you rather than routing directly to the next, cheaper alternative?

The answer may not be user experience. It could be liquidity, latency, settlement assurance, or something else.

At @bcap, we’re spending a lot of time thinking about this—whether in investment committee meetings or discussions with our engineering team. We don’t yet have a definitive answer. If you’re building products around Agents and have your own perspective on value capture in the Agent era, we’d love to hear from you.

Original link