While intent-centric services offer a wide range of user experience advantages, one only needs to look at taxis as an example to understand where the system might go wrong.

Providing detailed directions for all our taxi rides, similar to the traditional model of specifying each step in a blockchain transaction, would be cumbersome and error-prone.

However, the "trust the driver" approach also has its issues, which are more analogous to intent-centric systems: we’ve all had the experience of hopping into a taxi in an unfamiliar city, hoping for a quick ride, only to find ourselves sitting awkwardly as the driver takes a suspiciously long route, running up the meter.

In this analogy, the taxi driver is like a solver in an intent-centric system: trusting the solver to handle the task means trusting them to execute it honestly.

Intent-centric programs typically have systems in place to keep solvers honest, meaning a more accurate analogy might be Uber, which allows drivers to check pre-set pricing and in-app routing. But ride-sharing apps only further highlight the risks of intent-based systems: anyone who has experienced Uber’s price surges in recent years has seen firsthand how convenience can entrench large corporations at the expense of end users. The real risk of intent-centric systems isn’t just dishonesty but the potential emergence of new monopolies.

Paradigm, a prominent blockchain investor and researcher, highlighted these risks in a blog post: "While intents are an exciting new paradigm for transactions, their widespread adoption could accelerate a broader shift of user activity toward alternative mempools," Paradigm wrote. "If mismanaged, this shift risks entrenching centralization and rent-seeking intermediaries."

As we grow increasingly accustomed to relying on these third parties to fulfill user intents, these companies may begin acting in their own interests—either by charging higher fees (like Uber) or by filling orders in ways that serve them rather than users.

Although most intent-centric services outsource to competitive solver markets—ostensibly to avoid centralization—it’s still possible for some firms to dominate the space.

For example, one could imagine a cryptocurrency exchange building solvers to dominate "buy" and "sell" use cases—effectively funneling all market activity onto its own ledger. The exchange might initially subsidize fees as a way to crowd out competitors, but once it captures the market, it would raise prices.