Pioneer of the stablecoin. Co-founder of Tether.

The branch served one purpose: proximity. For two centuries, that was sufficient. The bank was where you went because the bank was the only place the functions existed. Proximity was scarce.

Proximity is not scarce. The smartphone reached more people faster than any previous technology in history, including electricity and running water. In most developing markets, smartphone penetration exceeds bank account penetration by a significant margin. The 1.4 billion adults globally who remain unbanked are not unbanked because they lack devices. They are unbanked because banking has always required documentation standards, minimum balances, and credit histories. Those requirements were always the source of banking's profit, not the service itself.

A wallet connected to this stack reaches anyone with a phone. No branch required. No documentation threshold. No minimum balance designed to exclude the people who would use the account the most and profit the issuer the least. The same infrastructure that serves an institutional investor running treasury operations serves a merchant in Nairobi running a micro-business on a feature phone.

Banking stops being a place you are admitted to. It becomes a function you carry with you.

A bank is four functions: custody, accounts, payments, yield. Branches existed because those functions required physical infrastructure. That constraint disappeared twenty years ago. What didn't disappear were the extraction mechanisms that grew up around the infrastructure. The settlement float generated by deliberately slow ACH windows. The fee income from geographic monopoly. The yield captured on deposits that holders never see.

The 1-to-3-day settlement window is not a technical limitation. The technology to move money in three seconds has existed for years. The window persists for two reasons. Float is profitable: every dollar in transit generates yield for the institution holding it, and that calculation is so deeply embedded in banking infrastructure that most people never notice it. And regulation, written for a world where settlement actually took days, has been slow to catch up to a world where it doesn't have to. The first is a revenue model. The second is a legacy frame that is finally starting to shift. Together they represent hundreds of billions of dollars annually transferred from account holders to the institutions that serve them.

WeFi is rebuilding the four functions of banking without the extraction mechanisms. Yield passes through to the holder as a structural feature, not a promotional rate. Settlement moves in seconds because there is no float to preserve. The assets never leave the wallets they came from, because WeFi is not a balance sheet. There are no pooled deposits to lend, no spread to capture, no extraction mechanism that depends on temporarily owning the customer's money. Banking becomes software that runs on any device, in any jurisdiction, without the gatekeeping infrastructure that was always the source of the profit, not the service.

In 2014, depositing one dollar and receiving one digital dollar in return was genuinely radical. The mechanism was auditable. Every promise was kept. Tether became one of the fastest-growing monetary systems in history precisely because the original contract was honored at scale.

What it didn't solve, what it couldn't solve within its own structure, was who benefited from the economics of that scale. When stablecoins reached hundreds of billions in circulation, the interest on the underlying reserves became a number that mattered. It flowed almost entirely to the issuer. The holders who provided the network effects. The merchants who gave the token its utility. The users who made the reserve valuable. They received none of it.

The second-generation model introduced by STBL separates what the first generation fused. The spendable token remains what it was: liquid, collateralized, redeemable at par. The yield from the underlying reserves becomes a separate asset. Transparent, programmable, transferable. And it flows to the people who created the reserves in the first place. Mint the stablecoin by depositing real assets, earn the yield those assets generate. The stablecoin becomes honest about its own economics for the first time.

Gold is not the most interesting asset. It is the most honest starting point.

There are no cash flows to model, no credit assumptions to verify, no jurisdiction to dispute. Five thousand years of track record as a store of value that no sovereign can inflate. Gold works as the foundation of a sound monetary system for exactly one reason: no one can make more of it. That constraint is why every government that has wanted to spend beyond its means has eventually moved off the gold standard. The constraint is the feature.

On-chain gold does something that physical gold and paper gold certificates never could: it enforces its own rules. Rehypothecation is one of the oldest extraction mechanisms in finance and one of the hardest to police. It works by pledging the same collateral multiple times in a chain of lending that obscures the true state of ownership. The 2008 crisis was partly a rehypothecation problem: the same assets had been pledged so many times through so many intermediaries that no one could reconstruct who actually owned what. Gold has lived with a quieter version of the same problem for decades. Paper gold certificates have historically traded at multiples of the physical gold actually held in vaults, which means most people who think they own gold own a claim on someone else's claim. A tokenized gold position carries its own provenance. The chain of custody is visible. The pledge history is on the ledger. The protocol can enforce a hard ceiling on how many times any unit of collateral can be re-pledged. Not because a compliance officer is checking. Because the rule is in the asset.

This matters for Islamic finance as much as it matters for institutional risk management. Shariah law has prohibited riba, gharar, and maysir for fourteen centuries. The prohibition on charging interest on money. The ban on selling what you do not own. The restriction on pure speculation. These are not arbitrary rules. They are a systematic effort to prevent extraction from the people participating in a financial system. The problem has always been enforcement: halal finance required separate products, separate compliance structures, separate institutions. Programmable assets collapse that requirement. The rules that make an asset Shariah-compliant are a configuration of the token, not a separate product category. The same infrastructure serves conventional and ethical financial systems because the ethics are encoded in the asset, not bolted on by an institution.

AUM begins with gold because trusted collateral is the cleanest proof that a real-world asset can become a productive system. A reserve asset that earned nothing becomes a working one.

Once collateral is programmable, the money issued against it has to become programmable too.

Every financial system begins with a ledger and a set of rules. What it also begins with, less visibly, is a separation: the rules live in one place, the assets in another. Laws, compliance departments, legal contracts, custodians. That separation is not just an architectural choice. It is the source of most financial extraction. The people who own the rules capture value from the people who hold the assets.

Programmable assets collapse that separation. The asset carries its own rules. Transfer conditions, lifecycle triggers, compliance requirements. All of it encoded in the token, enforced by the protocol, requiring no intermediary to police.

We proved this in one of the hardest environments first: global consumer brands. Through SmartMedia Technologies, programmable tokens were deployed to more than 11.5 million wallets across Visa, PepsiCo, Samsung, FIFA, and the Olympics. The model worked at the scale of the global consumer economy. Dual opens that infrastructure as a decentralized protocol any builder can use, so that assets anywhere can carry their own identity, logic, permissions, and lifecycle.

The rules used to live in institutions. Now they live in the assets.

Before money can be programmable, the object holding value has to be programmable first. Everything above this layer assumes it already exists.