Blockchain as the everything ledger

Blockchain as the Everything Ledger

I increasingly think blockchain may end up mattering for a reason that has very little to do with convincing humans to use crypto.

For years, the industry has asked some version of the same question: why would an ordinary person put something onchain when a bank, company or database already does the job?

It is a fair question.

But perhaps humans were never the most natural users of blockchains.

Machines might be.

Stablecoins are already turning money into software. A dollar on a blockchain is not just something you can spend. It is something software can hold, move, split, escrow and deploy according to rules.

That matters once software starts acting economically.

AI agents will increasingly buy compute, access data, call APIs, pay other agents and receive money for work. They will not want to navigate banking interfaces designed around human beings. They will want money they can interact with directly.

And once money becomes programmable, I suspect other forms of value start moving in the same direction.

Take Clubcard points.

They feel completely different from money, but structurally they are not that alien. They are an entry in Tesco’s database representing something I have earned and can later exchange for value.

The same is true of airline miles, gift-card balances, game currencies and plenty of other economic rights.

Even much of traditional finance is ultimately a question of ledgers. Who owns the share? Who holds the bond? Who owes the money?

The economy is already full of databases keeping score.

What changes if more of that value becomes programmable, transferable and readable by machines?

That is where I think blockchain starts to become more interesting.

Not as the database for everything, but as the ledger for economic state.

Who owns this?

Has this been paid for?

Is this agent authorised to spend?

Can this asset be transferred?

Has this contractual condition been met?

Today those answers live across thousands of private systems. In an economy increasingly populated by autonomous software, there is obvious value in having shared rails through which those systems can interact.

And AI could increase the scale of that economy enormously.

There are eight billion humans. There could eventually be vastly more agents.

One person may have dozens working on their behalf. A company could operate millions. Agents may hire agents, buy from agents and sell to agents.

They could also transact in quantities that make little sense for humans.

A fraction of a penny for a model call. Another fraction for a piece of data. Another for compute. Thousands of tiny exchanges taking place continuously in the background.

This is where something flips.

Many of the things that make crypto feel strange to humans are almost irrelevant to machines.

An AI does not care about a polished banking app. It does not need a card in its wallet. It does not find hexadecimal addresses ugly.

It cares about whether it can transact, whether ownership can be verified, whether the rules are programmable and whether the network will still be there when it needs it.

Perhaps blockchain has spent its first fifteen years being judged through human UX when its more important future use case is machine infrastructure.

That also raises the question of who gets to own the network.

If autonomous agents become major participants in the economy, it seems unlikely that the ideal settlement layer is one controlled by Google, Amazon, OpenAI or any other single company.

The more important the network becomes, the more valuable neutrality becomes.

That is why I keep coming back to Ethereum.

I do not know whether Ethereum wins. There will almost certainly be many chains, Layer 2s and specialised systems.

But Ethereum has spent more than a decade becoming something increasingly boring in the best possible way: open infrastructure that keeps running.

No company owns it. Anyone can build on it. Stablecoins and other digital assets already move through its ecosystem.

And if it succeeds, most people may never notice.

Nobody thinks about TCP/IP when they open Instagram.

In the same way, somebody in 2035 might ask an AI agent to book a holiday, negotiate a service or purchase some digital resource without having any idea that part of the transaction ultimately settled on Ethereum.

That is the version of blockchain I find increasingly compelling.

Not millions of people suddenly deciding they love crypto.

Billions, perhaps eventually far more, pieces of software quietly using blockchains because programmable money and neutral settlement simply make more sense for machines.

AI gives software intelligence.

Stablecoins give it money.

Blockchain gives it somewhere to keep score.

And that ledger may end up recording far more of the economy than we currently imagine.