From AI-Augmented Capitalism to the Race Between Frontier Creation and Rent Extraction

For 250 years, capitalism has gotten progressively better at exploiting what it already knows. AI may compress the useful life of some forms of advantage, but not necessarily the advantage of the firm itself.

Capitalism has repeatedly reinvented itself around new forms of abundance: steam, electricity, computing, the internet, cloud. Today we are entering another phase, AI-augmented capitalism, in which machine intelligence is being inserted into companies, jobs and institutions shaped by all those earlier ones.

Another way to read that history is as a steady improvement in exploitation – not in the political sense, but in the economic sense of scaling what we already know is valuable: making it bigger, cheaper, faster and more widely available.

Exploration is different. It means finding a new technology, an unmet need, a new market, or a problem framed differently enough that an opportunity becomes visible. We have become very good at systematizing exploitation. Exploration remains far less predictable.

I think AI might accelerate that asymmetry. Access to the underlying capability has spread unusually quickly: Bick, Blandin and Deming found overall adoption running ahead of comparable early adoption for PCs and the internet. But trying an AI product is not the same as rebuilding a bank, factory or hospital around it. Deep integration still runs into regulation, legacy systems, liability, physical infrastructure and organizational inertia.

I don’t think there is a clean historical clock showing business advantage collapsing from decades to years to months, and I don’t know of a convincing empirical mechanism that measures that. What is changing is narrower: reproducible capability can spread faster even when the economic advantage surrounding it does not.

I call this the exploitation half-life: the time between discovering a valuable capability and competitors acquiring enough of it to materially erode the excess returns it generates. I use the term as a diagnostic frame, not an empirical constant.

For much of capitalism’s history, successful exploration bought a long period of exploitation. If you discovered a better production system in 1910, competitors couldn’t reproduce it the following week. They had to understand it, raise capital, build factories and develop the operating capability required to compete. A successful discovery bought time, and time created rents.

AI may shorten that interval for some capabilities.

The Half-Life of Capability Is Not the Half-Life of the Firm

Let’s take a bank as an example to think through this.

Suppose it uses AI to cut the cost of processing a type of loan by 40 percent. Initially, much of that gain can appear as profit.

Then competitors begin catching up. Models spread, vendors package workflows, employees learn the practices. The productivity gain remains, but the bank’s ability to earn unusually high returns simply because it has that capability weakens.

That is the simple commoditization story we all know.

Imagine the bank gets two years of unusually high margins from its early advantage. It can use that surplus to acquire a fintech, lock in distribution, attract customers, deepen proprietary data, build infrastructure or strengthen its regulatory position. By the time competitors acquire similar AI capability, the original advantage may be gone. The assets financed by it may not be.

AI may shorten the half-life of reproducible capability without shortening the half-life of the firm’s overall advantage.

This is close to the argument I made in Migrating Scarcity: abundance does not eliminate scarcity; it moves it. As one capability becomes easier to reproduce, value migrates toward whatever remains difficult to copy, scale or obtain. AI doesn’t erase moats. It changes which moats matter.

There is also evidence that technological diffusion doesn’t automatically destroy rents. Autor, Dorn, Katz, Patterson and Van Reenen document rising sales concentration and the growth of highly productive “superstar firms” across many U.S. industries. De Loecker, Eeckhout and Unger document a substantial rise in average markups since 1980, driven disproportionately by firms at the top of the distribution.

That doesn’t contradict commoditization. It points to what firms can do with the gains before the capability spreads: convert a temporary technological advantage into a more durable position. The capability commoditizes. The scarcity it helped finance does not.

Now let’s take the bank one step further.

An AI-augmented bank asks how AI can make its existing business more efficient. An AI-native bank would assume that many reproducible operating advantages will keep getting cheaper and more widely available, and organize accordingly.

I am using AI-native differently from the increasingly common technical meaning focused on agentic software or autonomous workflows. I mean it at the level of strategy and capital allocation: an institution designed around the expectation that reproducible advantages will decay faster.

Such a bank might make more small bets, scale winners quickly, expect imitation earlier, retire commoditized advantages sooner and keep more capital and organizational capacity pointed toward the next opportunity. It might pursue markets yesterday’s economics made too small, uncertain or expensive to serve. No bank fully operates this way today. The important point is the assumption behind the design.

Once something works, AI helps scale it. Others learn. The reproducible part spreads. The firm then has two broad ways to sustain extraordinary returns: create something new, or convert temporary surplus into something harder to reproduce.

That is where the argument becomes less about AI and more about capitalism.

Frontier Creation and Rent Extraction

Capitalism needs rents. Without the prospect of unusually high returns, there is much less reason to take unusual risks. The important question is what those rents finance next.

One path is frontier creation. Surplus is recycled into new science, technologies, products, markets, infrastructure and institutions. New economic territory opens, new scarcity appears, and eventually some of it becomes reproducible too.

The other is rent extraction. Surplus is used primarily to acquire, consolidate or defend scarce assets that already exist: land, capital, compute, energy, networks, distribution, intellectual property, regulatory rights, proprietary data or physical infrastructure.

These aren’t two kinds of company but two uses of the same surplus, and the same company can do both. Scarce infrastructure or distribution can generate the rents that finance exploration, while successful exploration can create the next scarce asset. A data-center network or semiconductor plant may begin as frontier creation and later become a durable source of rents.

The second path has a self-reinforcing logic. Scarce assets produce rents; those rents can finance more scarce assets, stronger distribution, greater scale or a more defensible market position. The evidence on concentration and rising markups gives us reason to take that mechanism seriously.

AI could strengthen it. Small teams can now access analytical, technical and creative capacity that once required much larger organizations, which should lower the cost of creating new frontiers. But the same abundance makes capability a weaker place to hold economic power. If everyone gets better software but only a few firms control the distribution, data, energy, compute or regulatory access needed to turn it into economic scale, the scarce complement can become more valuable precisely as the software becomes abundant. The same pressure can operate inside a frontier attempt: as building becomes cheaper, more of the value can migrate toward scarce inputs such as foundation models, compute and energy.

That doesn’t mean concentration inevitably wins. New technology can destroy yesterday’s bottleneck, entrants can attack from another layer of the market, regulation can prevent temporary advantage from hardening into permanent control, and attractive new frontiers can pull capital toward building instead of defending. None of that happens automatically.

There is another constraint on frontier creation: demand. If AI eventually shifts a meaningful share of income from labor toward ownership, the economy can become extraordinarily productive while purchasing power becomes more concentrated. New frontiers still need customers. An economy that gets better at producing new things but worse at distributing the income required to buy them could tilt further toward asset ownership even if exploration remains vibrant.

The Direction of the Race

The future is therefore not a coin flip. Firms have a structural reason to convert temporary advantage into durable scarcity while they can, and existing scarcity can then generate cash flows that strengthen the position further. That gives rent extraction a head start.

AI could make frontier creation cheaper too. If small groups can experiment with capabilities that once required large organizations, the rate of frontier creation could rise alongside the rate at which known capabilities spread.

We don’t yet know which effect will dominate. But the burden is not symmetrical. If new frontiers aren’t created fast enough, capital doesn’t stop searching for returns. It moves toward owning and defending whatever scarcity remains.

That asymmetry is why AI-native capitalism is a more consequential idea than simply using more AI. Reproducible capability may keep getting cheaper while durable economic power migrates elsewhere. If so, the default pull is toward whatever remains scarce and ownable.

Frontier creation is what keeps that equilibrium from closing in on itself.

The defining economic question of AI-native capitalism may therefore be not whether we can create abundance, but whether the rents abundance creates are repeatedly recycled into opening the next frontier – or allowed to compound around the scarcity that remains.

A service starts timing out intermittently after a routine deployment. Nothing looks particularly alarming. The deployment itself appears clean, the underlying systems are healthy, and latency has increased but not enough to trigger the usual alarms.

An engineer eventually notices that the slower requests are interacting with a retry policy. The policy is perfectly sensible on its own. If a request takes too long, try again. But under the conditions created by the deployment, some requests are taking just long enough to trigger retries. Those retries create additional traffic, which increases latency, which causes more requests to retry. Neither the slower requests nor the retry policy is enough to explain the problem by itself. The problem is in the interaction.

Once the team understands what happened, it can fix it. Maybe the retry policy changes. Maybe the timeout does. Maybe the system gets better monitoring for this particular pattern. The exact engineering answer isn’t important here.

What interests me is what the company learns.

The most immediate lesson is very specific: change this retry policy so this particular sequence can’t happen again. That’s useful, and it should be done.

But somebody looking at the incident might extract a broader lesson. When latency starts rising, watch for retries that can amplify the problem. That lesson could help somewhere else in the system, even where the exact policy is different.

Go one level higher and the lesson becomes broader again. When individual components appear healthy but the system isn’t, look at the interactions between them.

All three lessons came from the same incident. All three are useful. But something changes as you move from one to the next. The first tells you almost exactly what to do, but only in a narrow situation. The second travels further, but somebody has to decide whether the conditions are similar enough for it to apply. The third could apply to almost anything, which is precisely why it tells you much less about where to look.

The more portable the lesson becomes, the more judgment it takes to use it.

I’ve been thinking about that because it feels like the problem sitting underneath the series of essays I’ve been writing about AI and work.

In The Average Is Lying to You, I argued that when AI takes routine work first, the average can stop describing the work people actually do. In The Side Door Is the Whole Building Now, I looked at what happens when exception processes built for occasional use start carrying much more of the human workload. The Dollar Threshold Is Static. Risk Isn’t. was about what happens to authority when the remaining work needs more judgment, while The Work Was Doing Two Jobs was about what happens when some of the routine work disappearing was also how people became experienced.

Then in The Normal Case Was the Company, I started wondering whether these were really separate problems. We didn’t simply build processes around normal work. Much of the organization itself was designed around it.

There is one more piece of that argument I hadn’t considered.

Organizations have always had to figure out how to make what one person learns useful to somebody else. One of the best ways we’ve found is to turn experience into something that travels. A rule. A process. A checklist. A standard. A precedent. A test.

When the lesson generalizes cleanly, this works extraordinarily well. The next person doesn’t need to repeat the original experience because the organization has already extracted what matters from it. We shouldn’t romanticize judgment and pretend everything needs to remain inside somebody’s head. Much of the progress of management has come from doing exactly the opposite.

The interesting cases are the ones where the lesson travels but doesn’t quite tell you what to do.

Take the retry incident. The company can preserve the exact technical fix. It can also preserve the broader heuristic about retry amplification. It can record the engineer’s reasoning, the alternatives she considered and the evidence that changed her mind. AI should make all of that much easier. Instead of reducing an incident to a few paragraphs in a postmortem that nobody reads six months later, an AI system could retain far more of what happened and retrieve it when a new problem looks similar.

That could be a significant improvement in how organizations learn.

But suppose six months later another service starts behaving strangely and the system retrieves this incident. Now there is a different question. Is the new problem actually similar to the old one?

Perhaps both involve rising latency, but for completely different reasons. Perhaps both involve retries, but the retry behavior that mattered last time is irrelevant this time. Perhaps the new incident looks nothing like the old one at the component level but has the same deeper structure: several things behaving normally and producing an abnormal result when combined.

Which similarity matters?

That’s judgment again.

And there is no reason to assume it always has to be human judgment. AI may become very good at this too. It may compare thousands of incidents, recognize structural similarities that people miss and tell an engineer which previous cases are genuinely relevant. It may eventually become better than most people at deciding which old lesson belongs with which new problem.

But then another distinction appears. Two situations may be similar in nine respects and different in one. Does the difference matter? A pattern may have predicted the right action fifty times. Is this the case where following it for the fifty-first time creates the mistake?

The problem isn’t that judgment can’t travel. Organizations have been making judgment travel for a very long time, and AI may allow much more of it to travel than ever before.

The problem is that every time judgment travels from one situation to another, somebody or something still has to decide where it applies.

That has made me think differently about the Context Loop I described in Migrating Scarcity. I was interested in getting relevant context back to the point where a decision is being made. I still think that’s important. But having more context available doesn’t eliminate the need to decide which context matters. In some situations, it may make that question more important.

This is the same scarcity movement at another level. Solve the narrow technical problem and value moves toward recognizing the pattern. Get better at recognizing patterns and value moves toward knowing which pattern applies. Automate that discrimination and the valuable question becomes which difference should override the pattern.

There may not be a permanent line somewhere in this chain marked “human judgment.” I don’t think the argument needs one. AI may keep moving further along it.

What matters is what happens to the work each time it does.

If AI increasingly takes the situations where we already know what matters, the remaining situations will contain a larger share where figuring out what matters is the work. If AI gets good at those too, attention moves to the next distinction it cannot yet make reliably.

That is why I’m increasingly skeptical of discussions about AI that end when a task has been automated.

The task was never necessarily the final constraint. It was simply where the constraint happened to be sitting at the time.

Organizations know how to make lessons travel. Rules do it. Standards do it. Precedent does it. Experienced people do it. AI may become the most powerful mechanism we’ve ever had for doing it.

But every time a lesson arrives somewhere new, there is another decision hiding inside the transfer: does it apply here?

Perhaps AI gets very good at that too.

And then the scarcity moves again.

Klarna is an interesting AI story partly because the AI worked.

When the company launched its AI customer-service assistant in 2024, the early numbers were extraordinary. After its first month, Klarna said the assistant was handling two-thirds of customer-service chats and doing work equivalent to 700 full-time agents. Average resolution time had fallen from 11 minutes to under two, repeat inquiries were down 25%, customer satisfaction was on par with human agents, and Klarna estimated the assistant would improve profits by $40 million in 2024.

A year later, something seemingly contradictory happened. Klarna began emphasizing humans again. CEO Sebastian Siemiatkowski acknowledged that the company’s aggressive focus on AI and cost reduction had affected service quality. The company began hiring again and put renewed emphasis on service quality and the role of people alongside AI.

It would be easy to tell this as another story about AI being overhyped, except Klarna didn’t abandon AI.

The more interesting question isn’t whether Klarna’s AI worked. It’s what happens to the human job when it does.

If machines handle a large share of straightforward customer interactions, the people who remain aren’t simply doing less customer service. They’re increasingly handling the things the automated system didn’t resolve, the things customers still want a person for, and the situations where something about the normal path didn’t work.

Klarna shows the first part of that pattern. The rest is what I’ve been trying to understand through this series. When machines disproportionately remove routine work, the averages can start lying to us about the work people actually do. As more human work consists of things that don’t fit the standard process, the side door can start becoming the whole building. When those decisions require more judgment, the dollar threshold starts looking increasingly crude because risk isn’t static. And when routine work was also how inexperienced people became experienced, the work was doing two jobs.

I’ve been treating those as different problems. I’m no longer sure they are.

The normal case wasn’t just most of the work. It was the organizing assumption of the company.

Think about how most organizations are built. Processes exist because enough situations repeat that we can decide in advance what should happen next. Jobs exist because enough tasks can sensibly be bundled together. Averages are useful because they tell us something about what people actually do. Approval limits work because fairly simple rules are good enough for a large number of reasonably predictable decisions. People get better partly because they do enough ordinary work before they’re asked to handle the difficult stuff. Even a cumbersome exception process makes sense when exceptions are actually exceptions.

The normal case made all of that possible, which is why I keep coming back to the argument behind Migrating Scarcity. When technology makes something abundant, scarcity doesn’t disappear. It moves to whatever the newly abundant capability still depends on.

There is something I didn’t take far enough in the book. What if the thing becoming abundant isn’t just a resource the organization uses? What if it’s the kind of work the organization was built around?

Much of the work companies are automating first has something important in common. It is structured enough, frequent enough or predictable enough to hand to a machine with confidence. Those are also many of the characteristics that made the work possible to standardize in the first place.

As more of that work moves to machines, what reaches people starts to look different. More of it involves ambiguity or disagreement. Facts don’t line up neatly. Objectives conflict. The policy covers most of the situation but not quite all of it.

And those aren’t just differences in difficulty. They put pressure on how work is measured, where decisions get made, when something gets escalated and how people learn to exercise judgment.

That raises a question I hadn’t really considered before writing these essays. If you were designing an organization mainly to handle that work, would you design the organization we have today?

I doubt it.

Take a different, hypothetical company. It automates 70% of a process. Unit costs fall, response times improve, customers get answers faster, and the business case delivers exactly what management promised. The automation was successful.

But the remaining 30% may now need different skills, different measures of productivity and more room for judgment. Escalation paths designed for occasional exceptions may start carrying much more of the human workload. Authority may still be calibrated to the old mix of work. And the company may eventually discover that some of the routine work it removed was also helping develop the people it now needs most.

The company solved the problem it set out to solve, but in doing so it exposed the next one. That’s the part that connects back to Migrating Scarcity. Removing a constraint and capturing the value released by removing it are not the same thing. If the constraint moves, eventually the organization has to move with it.

I think this may explain some of the frustration we’ll see as AI deployments mature. Automation rates will rise and unit costs will fall. Individual AI projects will show perfectly respectable returns. Yet executives may still wonder why the organization isn’t becoming as fast, adaptive or productive as the technology seemed to promise. The natural reaction will be to ask what else can be automated, and sometimes that will be exactly the right question.

But sometimes the AI will already have done its job. It will have removed enough of the normal work to expose something we hadn’t thought much about: much of the organization was built on the assumption that people would still be doing that work.

At that point, another model isn’t necessarily the answer.

The organization has to move with the scarcity.

For more than a century, one of management’s great achievements was learning how to make ordinary work repeatable enough to scale. We broke jobs into pieces, standardized processes and built organizations capable of handling enormous amounts of recurring activity efficiently.

AI may turn out to be extraordinarily good at precisely the work we became extraordinarily good at organizing. That doesn’t make the organization unnecessary. It changes what we need the organization to be good at.

We designed the organization around the normal case and built special machinery at its edges for everything that didn’t fit. AI may be taking the normal case first.

What happens when the edges are increasingly where the organization lives?