Why Evan Cheng Believes Most Blockchains Were Built on the Wrong Architecture, and What Sui Does Differently

In this episode of When Shift Happens, I sit down with Evan Cheng to discuss why he believes the architecture behind most major blockchains is fundamentally insufficient, what Sui was built to solve instead, and how the next wave of adoption may come. What makes the conversation especially interesting is that Evan is not simply arguing that Sui is faster or cheaper. He is making a deeper claim: that much of crypto is still built on a mental model of assets and ledgers that works for a narrow class of problems, but breaks down when the world gets more complex.

From Libra To Starting Over

Before co-founding Mysten Labs, Evan helped lead Facebook’s Libra, later Diem, effort. That project gave him direct exposure to what a blockchain optimised for payments could look like. But instead of taking that work, open-sourcing it, and launching a slightly modified version of it, he and his co-founders chose to start again.

As he puts it, when they considered simply turning Diem into a permissionless chain, they looked at each other and laughed. “No, I mean what’s the point of doing that?” That line captures the heart of the episode. Evan is not presenting Sui as a continuation of what came before. He left because he believed that even the underlying design they had worked on with Libra was too limited for the future they now saw coming.

Why He Thinks The Old Model Breaks Down

Evan’s core argument is surprisingly simple when stripped of technical language. He says that most major layer 1 blockchains are built around the idea of a ledger that tracks balances moving between accounts. That model works well when assets are basically the same. Money is the clearest example. If one wallet has ten units and another has five, the system only needs to record quantities moving around.

But Evan argues that this is not how the real world works. In the real world, assets change. They evolve through interaction. They acquire history, new states, new rights, new restrictions, and new meaning.

He uses baseball cards as an example. One hundred digital baseball cards may start out identical. But if one card is later signed by the player, it has changed. Its value, status, and identity are no longer the same as the others. The same logic applies to contracts, documents, financial instruments, and any digital object that can be altered through interaction with people, rules, or other assets.

In his view, blockchains designed mainly as balance sheets are trying to stretch themselves into handling a world of changing, composable assets that they were never built to represent cleanly in the first place.

The Case For Sui’s Object-Based Approach

This is where Sui’s difference comes in. Evan repeatedly returns to the idea that Sui is built around objects rather than just accounts and balances. The implication is that the chain is designed to model assets that are configurable, composable, partially private, subject to rules, and able to settle across more complicated workflows.

For Evan, this is also why “one-size-fits-all” approaches fail. He argues that the world does not divide neatly into fully public systems on one side and fully private systems on the other. Institutions may need some information to remain visible only to selected counterparties, while the outcome of a transaction still needs to settle on-chain and become usable in a broader environment. That middle ground, he suggests, is where existing architectures begin to struggle.

Product-Market Fit, But Only At The First Layer

One of the more thoughtful parts of the episode is Evan’s answer to the question of product-market fit. He does not argue that blockchain has already been fully proven, nor that Sui has conclusively won. In fact, he says flatly, “nobody has proven anything” at the level of mass adoption.

What he does say is narrower and more credible. Sui, in his view, has shown product-market fit at the infrastructure layer. The proof is that developers can build quickly on it, that a range of applications are emerging, and that parts of the stack are already being used in ways that demonstrate real technical capability. In other words, the rails are working. The platform has proven it can support meaningful experimentation, especially in DeFi and adjacent areas. 

What remains unproven is whether that can scale into the larger, more complex adoption wave he is targeting. 

Where The Next Adoption Wave May Come From

Evan is also clear that crypto-native trading and meme coin culture are not enough to sustain the ecosystem at the level many once imagined. He notes that speculation multiplied, but participation and liquidity did not expand proportionally. The industry became louder without necessarily becoming broader.

So where does he think the next real change comes from? Institutions, tokenised assets, stablecoins, and more efficient financial rails. He points to something as seemingly small as moving from T+1 settlement speed to T+0 settlement as potentially enormous. In traditional finance, transactions often follow a T+1 model, meaning a trade may happen today but the final settlement, when ownership and funds actually transfer, occurs the next day. Moving to T+0 settlement means that trade and settlement happen instantly. 

On-chain systems make this possible because the same infrastructure records the transaction and finalises it in real time. The result is faster movement of capital, fewer intermediaries, and significantly lower counterparty risk. This could dramatically increase the velocity of money, making financial markets more efficient and opening the door to entirely new kinds of financial products built directly on blockchain rails. Faster settlement increases the velocity of money, changes how efficiently capital can move, and opens the door to products that feel familiar to users but work better underneath.

Better Technology Is Never Enough

Still, one of the strongest insights in the conversation is that Evan does not romanticise technical superiority. He says plainly, “Just building a better technology is never enough.” People need education. Regulation, distribution, and mindshare all matter. Inferior technology can still win if it has better positioning, better awareness, or an earlier lead that proves hard to dislodge.

Evan Cheng is not just building a blockchain he thinks is better. He is building for a world he believes is coming, and betting that when digital assets become more complex, interactive, and deeply embedded in real systems, the gap between architectures built for that world and architectures patched together for yesterday’s problems may become impossible to ignore.

👉If you enjoyed reading the summary, head over to When Shift Happens on YouTube or your favorite podcast platform to access the full convo.

https://www.youtube.com/watch?v=5Tunu3t7kQ4