We've Been Asking the Wrong Question About Bitcoin

By Rob Frye

AI could make intelligence, labor and production dramatically cheaper. In that world, Bitcoin's most interesting feature may not be its price. It may be the one thing technology can't manufacture more of.

Bitcoin recently pushed to an eight-month high near $87,400 before falling back toward $83,000, and the conversation immediately returned to the usual checklist: interest rates, Treasury yields, the dollar, liquidity. Risk-on. Risk-off.

And that's fair. Bitcoin is volatile, liquid and perfectly capable of selling off with technology stocks when investors need cash.

I just think we're asking the right question on the wrong time scale.

Whether Bitcoin trades like a risk asset this week matters if you're trading this week. What happens to money if artificial intelligence changes the economics of labor, production and scarcity over the next 10 or 20 years is a very different question.

And that experiment has already started.

The world is getting very good at making things

Imagine intelligence becoming incredibly cheap.

Software that once took a team of engineers months to build gets produced in days or hours. Robots manufacture goods around the clock. Drug discovery speeds up. A company that once needed 5,000 employees might eventually produce twice as much with 500.

A lot of that sounds great. It probably will be.

But there's an awkward problem buried inside the abundance story: machines can make the products. They don't automatically create customers with money to buy them.

That's where this stops being just an AI story and starts becoming a monetary one.

The balance sheet we're bringing with us

Before getting too futuristic, consider the balance sheet we're taking into this new world.

The Congressional Budget Office estimates that the U.S. federal deficit reached roughly $2 trillion during the first 11 months of fiscal 2026.

Its longer-term baseline isn't exactly comforting either. Federal debt held by the public is projected to rise from about 101% of GDP in 2026 to 120% by 2036 - above the previous record of 106% reached just after World War II.

That doesn't mean America is about to default, and it certainly doesn't mean the dollar disappears.

It means we're walking into what could be one of the largest technological transformations in history without a lot of fiscal room for error.

Then AI potentially changes the other side of the ledger: taxation.

AI doesn't need to take everybody's job

The extreme AI story is easy to dismiss. Robots arrive, everyone loses their job and humanity spends Tuesday afternoon wondering what to do next.

We don't need anything close to that for the economics to change.

The World Economic Forum projects that labor-market transformation could create roughly 170 million jobs while displacing 92 million by 2030. People understandably focus on the net result: 78 million more jobs.

I keep coming back to the other number.

Ninety-two million jobs displaced.

That's an extraordinary amount of economic rearrangement in a short period.

More importantly, the number of jobs may not be the real issue. What matters is how much economic output eventually flows to labor compared with capital.

A business could conceivably double its output without doubling its payroll. Society gets more stuff, productivity rises and the owners of the productive assets do very well.

But the relationship between production and wages has changed.

Modern governments depend heavily on that relationship. People work. People get paid. Governments tax some of that income, while workers and employers fund employment-linked programs.

Start changing the first two steps and this is no longer just a debate about whether an AI model replaces an accountant. It's a question about the architecture underneath the tax system.

IMF researchers have examined exactly this problem, including how AI-driven labor disruption and changes in the distribution of income could require changes to tax and social-protection systems. Importantly, that's scenario analysis, not a prediction of mass unemployment.

No robot apocalypse required.

A meaningful shift in where the income goes is enough.

The sandwich factory problem

Here's the simplest way I know to explain what bothers me about the abundance argument.

Imagine somebody builds a fully autonomous sandwich factory.

AI buys the ingredients. Robots unload them, make the sandwiches, package everything, clean the factory and arrange delivery.

Instead of employing 1,000 people, the operation needs 30.

It produces 10,000 sandwiches an hour.

We've achieved sandwich abundance. Fantastic.

Except for one detail.

Who buys all the sandwiches?

The people who used to work there still need lunch. Producing something and distributing the purchasing power required to consume it are two completely different problems.

If automation pushes this dynamic far enough, society eventually needs a bridge between the two.

Maybe that's some version of basic income. Maybe citizen dividends. Maybe people own much larger shares of productive capital. Maybe governments tax capital or consumption differently.

I don't know which answer wins. I doubt there will be one clean answer.

But nearly every proposed solution eventually runs into the same problem:

Where does the purchasing power come from?

If labor supplies a smaller share of taxable income while governments are simultaneously expected to support more people through the transition, the pressure doesn't disappear. It moves - toward capital taxes, consumption taxes, borrowing, spending choices and, potentially, monetary policy.

And remember the balance sheet we're starting with.

That's the connection I think gets missed.

The abundance story doesn't just change what gets made. It changes the pressure placed on money and the institutions behind it.

This is where 21 million gets interesting

Bitcoin's price isn't stable.

Its monetary rule is.

Under Bitcoin's consensus rules, issuance declines on a predetermined schedule and total supply approaches 21 million coins.

AI can create another image. Another song. Another piece of software.

Eventually, robots may manufacture another car, phone or house using a fraction of today's human labor.

AI cannot simply wake up tomorrow morning and decide the economy would function better with 31 million Bitcoin.

Changing that requires convincing the network to accept a different rule.

Now, there's an obvious caveat.

Scarcity by itself is worthless.

I could create a token this afternoon with a maximum supply of seven. Nobody has to care.

Scarcity matters when it meets demand.

And that's what makes the next part of this experiment so interesting.

Suppose AI makes intelligence cheaper. Robotics makes labor cheaper. Automation makes manufacturing cheaper. Digital content becomes almost infinitely reproducible.

What remains difficult to reproduce?

Land. Energy. Certain natural resources. Human attention. Trust. Established networks.

And perhaps monetary assets whose supply constraints people actually believe.

The more technology makes abundant, the more value may migrate toward things technology can't simply multiply.

That doesn't prove Bitcoin wins.

It does make the economic value of credible scarcity worth taking seriously.

A view from the trenches

I've spent a good part of my career watching crypto markets from the inside, and one thing becomes obvious pretty quickly: a predictable protocol doesn't produce a perfectly efficient market.

Through my current work with Monivo, we've now measured more than 900,000 provider quotes in the Monivo Crypto Swap Rate Index. Across comparable quotes for the same transaction, the median gap between the best and worst recorded payout has been about 2%.

The dataset isn't the entire market and quotes aren't completed trades, so I wouldn't pretend it proves more than it does.

But I like what it illustrates: even when an asset's supply rule is predictable, price discovery around it can remain surprisingly messy.

Protocol certainty and market efficiency are not the same thing.

What would prove this wrong?

Bitcoin commentary has an annoying habit of turning every development on Earth into another reason Bitcoin must go up.

Inflation? Bullish.

Deflation? Somehow bullish too.

Strong economy, weak economy, rate cut, rate hike - give people enough time and somebody will explain why all of them are good for Bitcoin.

That's not analysis.

There are plenty of ways the argument I'm making could be wrong.

AI could generate new categories of human work faster than it displaces old ones. Productivity gains could flow broadly through higher wages and widespread ownership rather than concentrating heavily among owners of capital. Governments could successfully redesign their tax systems. Fiscal conditions could improve.

And even if credible scarcity becomes more valuable, there's no law saying humanity has to choose Bitcoin.

People may choose equities. Real estate. Gold. Energy infrastructure. Land. Or something that hasn't been invented yet.

Bitcoin carries its own risks too: regulation, technological competition, extreme volatility, changing market structure and the simplest risk of all - future demand may not develop the way Bitcoin holders expect.

Twenty-one million is a supply rule.

It isn't a promise about price.

Saying that out loud makes the argument stronger, not weaker.

So what does the rally actually tell us?

Bitcoin's recent move is encouraging for the industry, but I wouldn't use a rally as proof of this thesis any more than I'd use the next 20% drawdown to disprove it.

Gold can fall during a liquidity crisis too. An investor who desperately needs dollars sells what can be sold. That doesn't necessarily tell us what he thinks about the asset's role over the next 20 years.

Price behavior during a panic and economic purpose across a generation are different questions.

Bitcoin can behave like a high-beta technology stock on Wednesday afternoon while also representing an attempt to insure against a different set of risks measured over decades: fiscal stress, monetary discretion, changes in the tax base and uncertainty over who ultimately captures the productivity gains AI may create.

That's a much bigger discussion than whether the next central-bank move is 25 basis points.

Central banks can change the weather.

They have a much harder time redesigning the climate.

The question I can't shake

We're entering one of the strangest economic experiments of our lifetimes.

Technology may make intelligence less scarce. Automation may make human labor less central to production. Robotics may make physical goods dramatically cheaper to manufacture.

At the same time, governments are carrying enormous debts and may eventually have to rethink systems built around employment, wages and the taxes attached to them.

Maybe the transition goes beautifully.

I hope it does.

Maybe AI creates so much new wealth and so many new kinds of work that today's concerns look quaint 20 years from now.

I hope that happens too.

But perhaps we've been asking the Bitcoin question backwards.

Instead of asking why anyone would want a deliberately scarce digital asset in an age of technological abundance, maybe the better question is:

Why do we assume credible scarcity becomes less valuable when almost everything else becomes easier to create?

About the author: Rob Frye founded cryptocurrency exchange Xcoins in 2016 and later served as CEO as the company established operations in Malta and became one of the early platforms to receive a Class 3 VFA license from the Malta Financial Services Authority. He is currently involved in strategy and product development at Monivo, a non-custodial crypto swap aggregator, and has a financial interest in the company. The views expressed are his own and do not constitute investment advice.