Stop #288 - The Last Scarcity

Thursday, February 26, 2026, may be remembered as the day the future stopped being a hypothesis.

Jack Dorsey, co-founder of Jack Dorsey and CEO of Block — the company that controls Square, Cash App, Afterpay, and the streaming platform Tidal — announced the layoff of 40% of the company’s workforce. More than 4,000 people out of 10,000. Not gradually, as other technology companies have done over the past few years. All at once.

The peculiarity of this drastic cut is that it does not stem from financial difficulties. On the contrary: Block’s gross profit grew by 24% over the past year, reaching $10.36 billion. In December 2025 alone, the company generated $1 billion in gross profit. The business, as Dorsey wrote in his letter to shareholders, “is strong.”

The stated reason is different, and Dorsey didn’t mince words: “Something has changed. we’re already seeing that the tools we’re building and using, combined with smaller and flatter teams, are enabling a new way of working that radically changes what it means to build and run a company. And this is accelerating rapidly.”

The company now aims to generate $2 million in gross profit per employee - four times the pre-Covid efficiency. We are witnessing the admission that much of the work performed by humans until yesterday can be done by machines today.

This isn’t the first mass layoff attributed to AI. Pinterest, CrowdStrike, Chegg, and dozens of other companies have made similar announcements in recent months. But Block’s is different. It’s the first where a CEO of a major publicly traded tech company declares it openly, without hiding behind euphemisms like “restructuring”, “resource optimization” or “adapting to market conditions”.

Block’s stock rose 24% in after-hours trading. Wall Street rewarded efficiency.

Emad Mostaque, founder of Stability AI and author of the recent essay The Last Economy, describes the economic history of humanity as a series of “inversions” - moments when the fundamental source of economic value flipped.

According to him, the first inversion occurred when land ceased to be everything. For 10,000 years, controlling land meant controlling the world. The Pharaohs measured their power in alluvial plains, the Romans in wheat fields. Then, around 1780, James Watt perfected the steam engine. Within decades, a brilliant landless merchant was more powerful than a foolish king with fertile fields. Value had shifted from what you owned to what you could organize.

The second inversion came when hands became obsolete. The era of labor dominance lasted two centuries, long enough to seem permanent. Unions, rights, the eight-hour workday. But in 1947, Bell Labs invented the transistor. By 1980, a computer could manage a production line. In 1920, Ford’s River Rouge plant employed 100,000 workers. By 1990, it produced more cars with 5,000. Value had shifted from labor to capital.

The third inversion was subtler: capital itself became ephemeral. In 1998, Kodak employed 170,000 people and was worth $31 billion. In 2012, Facebook bought Instagram for one billion. Instagram had 13 employees. No factories, no inventory, no physical products. Just the ability to organize human attention at scale. The entire photography industry evaporated in less than a decade. Not because people stopped taking photos, but because they started taking infinite ones. The scarcity that gave photos value vanished.

Now we’re at the fourth inversion. The final one. The one from which there’s no return.

On November 30, 2022, OpenAI released ChatGPT to the public. One million users in five days. One hundred million in two months. The fastest adoption of any technology in human history. But speed isn’t what matters. What matters is the meaning. Throughout all of human history, intelligence was a form of labor - scarce and locked inside human skulls. Now, for the first time, intelligence has become a form of capital. It can be copied infinitely. It improves recursively.

hatGPT to the public. One million users in five days. One hundred million in two months. The fastest adoption of any technology in human history. But speed isn’t what matters. What matters is the meaning. Throughout all of human history, intelligence was a form of labor - scarce and locked inside human skulls. Now, for the first time, intelligence has become a form of capital. It can be copied infinitely. It improves recursively.

At the beginning of 2024, an average American writer earned about 45 dollars per hour. By March 2024, API access to large language models allowed generating a 750-word draft for about six cents.

In 2020, Canadian entrepreneur Jeff Booth published The Price of Tomorrow: Why Deflation is the Key to an Abundant Future. The central thesis is very simple: technology is deflationary by nature. Every innovation reduces the cost of producing goods and services.

Booth provided concrete examples. The cost of a flat-screen TV has dropped 90% in twenty years while quality has increased exponentially. The cost of sequencing a human genome went from $3 billion in 2003 to less than $1,000 today. Digital storage - hard drives, SSDs, etc. - costs today an infinitesimal fraction of what it cost in the ‘90s.

The problem, according to Booth, is that we’ve built an economic system that requires perpetual inflation to function. Central banks create money from debt. Over 90% of money in circulation is born when commercial banks issue loans. Debt requires growth to be repaid. Growth requires prices to rise, not fall. But technology pushes in the opposite direction.

Now artificial intelligence is accelerating this process at exponential speed. The marginal cost of intelligence - that resource that for millennia was the exclusive monopoly of the human brain - is trending toward zero.

Mostaque calls this phenomenon the “Abundance Trap“. We’re about to reach post-scarcity in the realm of intelligence, and our scarcity-based economic system will process this abundance as poverty.

The paradox is evident in the numbers. Stock markets are at all-time highs. GDP is growing. Official unemployment is low. By every metric our leaders watch, we’ve never been so prosperous. But life satisfaction is at historic lows. “Deaths of despair“ - suicides, overdoses, alcoholism - are at epidemic levels. An entire generation cannot afford a home or start a family. We are richer on paper and poorer in spirit than we’ve ever been in modern history. On this topic, I highly recommend Jack Mallers’ speech, CEO of Strike, at the BTCPrague 2025 conference.

This disconnection, Mostaque writes, “is the first siren of a collapsing paradigm.* We’ve built a civilization so perfectly backwards that our greatest triumph is becoming our extinction event.*”

There’s a concept that Mostaque introduces in his essay that deserves particular attention: the “Metabolic Rift“.

For 10,000 years, all economic work was performed by metabolic engines called human beings. We need food, shelter, rest. We need a complex social structure to function: families, communities, institutions. The entire superstructure of wages, prices, and money was built on this non-negotiable thermodynamic foundation. Capital needed labor. Labor needed calories. This was the great pact.

Artificial intelligence breaks this pact. An AI doesn’t need sustenance. A robot doesn’t need shelter. They don’t need vacations, healthcare, continuous training, motivation. They don’t form factions in the office. They don’t interpret instructions as personal criticism. They don’t leave the company for better offers. They only need electricity.

We have created a form of work that Mostaque calls “*non-metabolic labor*“.**

Human labor, with its enormous overhead of biology and culture, cannot compete on price with silicon labor, which has none. When the marginal cost of cognitive work approaches the price of electricity, the value of a human mind - in purely economic terms - collapses. When land became less important than labor, humans became workers. When labor became less important than capital, humans became knowledge workers, using their intelligence. But when human intelligence becomes less important than AI, humans become… what?

One might think to answer: “AI trainers and supervisors“. But this is transitional at best. AI systems are already training other AI systems. Human supervision is a bottleneck the system is actively trying to eliminate.

The honest answer is: we don’t know. For the first time in economic history, we face an inversion without an obvious landing place. We are the generation that will live through the discontinuity. The last humans to remember when human thought had economic value. The first to discover what comes after.

Let’s return to Jeff Booth. His analysis started from a simple observation: if technology constantly reduces production costs, prices should fall. In a free market, technological deflation would translate into lower prices for consumers. We would produce more with fewer resources, and the benefit would spread to everyone.

But that’s not what happened. Prices rose. Houses cost more, not less, despite improved construction technologies. Education costs more, not less, despite knowledge being more accessible than ever. Healthcare costs more, not less, despite medical advances.

The reason is that central banks have systematically offset technological deflation with monetary inflation. Every time technology reduced costs, the central bank printed enough money to raise nominal prices. The result has been a wealth transfer: the value created by technology has been captured by those who own financial assets, not by those who produce or consume.

Artificial intelligence brings this paradox to breaking point. If machines can do almost all cognitive work, productivity will grow exponentially. But in the current system, this will translate into mass unemployment and further wealth concentration.

In a world where almost nothing is authentically scarce, what will still have value?

Let’s run through the options.

Gold

Its limited supply on Earth made it the natural basis for money. But gold is scarce only on our planet. The asteroids in our solar system contain quantities of precious metals that would crash any terrestrial price. The asteroid 16 Psyche, already the target of a NASA mission, contains iron, nickel, and gold with an estimated value of $10,000 quadrillion - more than all the world’s GDP produced in human history. SpaceX is reducing the cost of space access at dizzying speed. Asteroid mining isn’t science fiction. Gold’s scarcity is a question of “when,” not “if.”

Real Estate

Here the situation is more nuanced. Land is effectively limited, especially in desirable locations. But the value of real estate isn’t just land - it’s construction, materials, labor, energy. If robots can build houses without human supervision, if materials can be produced at decreasing marginal cost, if energy becomes practically free thanks to fusion or advanced solar, the cost of housing will tend to fall. What will remain is the value of location - proximity to urban centers, natural beauty, desirable communities. But this is a value of relative scarcity, not absolute.

Energy

Wars fought over oil. Economies built around access to energy resources. But solar is reaching costs so low as to make energy practically free in many regions. Nuclear fusion, long promised and never achieved, is finally showing concrete progress. In a future of abundant energy, what remains scarce?

Amid this landscape of evaporating scarcities, there’s an anomaly. Something that doesn’t follow the rules. Something that remains irreducibly scarce even in a world of total abundance. You’ve understood what I’m referring to.

Mostaque, in his essay, criticizes Bitcoin. He writes that its Proof of Work mechanism “consumes the energy of a nation just to protect its ledger“ and that its “deflationary nature incentivizes hoarding rather than circulation, starving the system of the very interactions that build intelligence and network capital“.

It’s a common criticism and - as often happens when it comes to Bitcoin - groundless. As they say: nobody’s perfect, not even Mostaque.

Bitcoin is a form of mathematical scarcity. It is encoded in the protocol. It cannot be altered without the consent of a distributed network of independent nodes that have no incentive to alter it. The 21 million bitcoin that will ever exist are not a promise made by a company or a government. They are a mathematical property of the system, verifiable by anyone.

We are talking about the only form of scarcity that is immune to technological progress.

It is the exact opposite of everything else. The cost of almost everything tends toward zero with sufficient technological progress, but the marginal cost of producing a bitcoin cannot tend toward zero—not even if ASICs were to become exponentially more efficient and, at the same time, cheaper. The difficulty adjustment mechanism would keep the supply fixed.

In a future where technology makes everything abundant, Bitcoin could represent the only anchor—the only mathematical certainty in a world of runaway variables.

In his book, Emad Mostaque asks: “Now that you no longer need to do anything to survive, what will you choose to be?”

It is a compelling question, but perhaps there is a more urgent one.

In a world where almost everything is becoming abundant, where will you store the fruit of your choices?

The window to position yourself is becoming increasingly narrow, because the world we once knew is ending, and the one that is coming will operate under different rules.