When Every Industry Starts Looking Like Cloud Computing with a Balance Sheet

Hatched by Ben H.

May 02, 2026

9 min read

7 views

72%

The strange convergence nobody should ignore

What do a health insurer and an artificial intelligence startup have in common?

At first glance, almost nothing. One sells risk management, claims processing, and healthcare administration. The other sells machine intelligence. Yet both are increasingly shaped by the same underlying force: who controls the infrastructure that makes everyone else function.

That is the real story hiding in plain sight. The most important companies of the next decade may not be the ones with the flashiest consumer products or the loudest brand narratives. They may be the ones that own the pipes, the platforms, the data, the distribution, and the leverage points that turn complexity into recurring revenue.

A health giant reporting massive financial strength and a cloud titan investing billions into an AI model builder are not separate stories. They are evidence of a single economic pattern: modern value accrues to whoever can become indispensable inside someone else’s workflow.

The new monopoly is not always a monopoly on customers. It is often a monopoly on coordination.

That idea connects healthcare, cloud computing, and frontier AI in a way that is easy to miss if you only look at the surface. But once you see it, a lot of corporate behavior suddenly makes sense.

From products to position: the real asset is embeddedness

The old economy rewarded companies for making better products. The newer economy increasingly rewards companies for becoming the place where other people must operate.

Think about the difference between selling a seat on a plane and owning the airport. The airline can earn money, but the airport controls traffic, access, scheduling, and a thousand small dependencies that make the whole system possible. In digital markets, cloud platforms are airports. In healthcare, claims systems and payer networks are airports. In AI, model providers and infrastructure partners are becoming airports.

This is why financial results alone can be misleading. A firm can post extraordinary numbers and still be revealing something deeper than profit. Strong financial performance can signal that it has achieved structural embeddedness, meaning other organizations depend on it so heavily that replacing it becomes impractical.

That is especially true in industries with high friction. Healthcare is full of regulations, legacy systems, billing complexity, fragmented stakeholders, and operational inertia. AI is full of compute bottlenecks, model training costs, safety concerns, deployment complexity, and the need for trust. In both worlds, the winning move is not just to invent something new. It is to become the layer that everyone else must pass through.

This is why the biggest strategic battles today are often not about products in the ordinary sense. They are about where the workflow begins, where it ends, and who gets to sit in the middle.

The new moat is not technology alone, but distribution plus dependency

It is tempting to describe the AI race as a contest of models. That is only half true. Models matter, but model quality is rapidly becoming one ingredient in a broader stack that includes compute, cloud relationships, chips, enterprise trust, and distribution.

Amazon’s large investment in Anthropic makes this obvious. The logic is not simply “we want a better chatbot.” The deeper logic is: if AI becomes a core interface for knowledge work, then whoever supplies the compute, the cloud, the deployment environment, and the embedded enterprise relationships gets an enormous strategic advantage.

Anthropic, for its part, is not just selling a model. It is selling a promise: safer output, longer context windows, more reliability on business and legal documents. In other words, it is trying to become the model enterprises trust when the cost of being wrong is high. That trust is not a nice-to-have. It is the moat.

This is where the analogy to healthcare becomes powerful. In healthcare, the product is rarely just medicine or care coordination. It is also reduction of uncertainty. Payers and large healthcare organizations win when they can predict, manage, and absorb risk better than fragmented competitors. Their strength comes from being the system that makes uncertainty legible.

AI companies are headed toward the same endgame. The most valuable models will not merely be smart. They will be the ones businesses can rely on when stakes are real, outputs need auditability, and mistakes carry legal or financial consequences.

So the moat is not just technical excellence. It is trusted dependency at scale.

In the platform era, the most valuable thing a company can be is hard to replace.

That sounds simple, but it has a profound implication: many so-called competitors are actually building adjacent dependencies inside the same ecosystem.

Amazon and Google can both invest in the same AI company because the prize is not exclusivity in the old sense. The prize is influence over the layer that will sit beneath many future applications. Likewise, a massive healthcare company can thrive not only by serving patients, but by controlling the administrative and financial rails through which care is paid for and coordinated.

Why scale keeps concentrating in the middle layer

There is a recurring shape to modern markets. The richest rewards often go not to the end consumer brand, and not even to the pure innovator, but to the middle layer that sits between fragmented demand and fragmented supply.

This middle layer is where the hard work of standardization happens.

Consider a simple example: a restaurant app is valuable, but the payment processor, cloud provider, and delivery logistics platform often capture more durable economics because they are used across many businesses. They do not depend on one customer’s taste. They become a utility embedded in many transactions.

Healthcare works the same way. A company that can standardize administration across a huge base of members and providers can gain a compounding advantage. Every new participant increases the value of the network and the cost of leaving it. The system becomes a kind of institutional gravity well.

AI is now entering that same structure. A model provider may start as a product company, but if it becomes integrated into cloud services, developer tooling, enterprise applications, and chip supply chains, it starts to look less like a software vendor and more like a coordination layer for cognition.

That phrase matters. We are not just automating tasks. We are building infrastructure for decisions, communication, drafting, analysis, and knowledge work. Whoever owns the interface to that cognition layer can shape how work gets done, what gets measured, and what gets normalized.

This is why the strategic competition around AI looks so intense. The companies involved are not merely betting on a product market. They are positioning for control over the next general purpose business utility.

And once a technology becomes a utility, the economics change. Margins can compress in the application layer, but the infrastructure layer often becomes more powerful, not less. The cloud wins. The railroads of the digital age win. The companies that own the training, the deployment, the distribution, or the proprietary workflows win.

The deeper lesson is uncomfortable: innovation does not always democratize power. Often it reorganizes it.

A framework for seeing the future: the dependency ladder

To understand why some companies become extraordinarily valuable while others merely participate in growth, use this simple mental model: the dependency ladder.

Every business sits somewhere on a ladder of how deeply others rely on it.

Optional: nice to have, easy to switch away from.
Useful: improves efficiency, but not mission critical.
Integrated: embedded in daily operations.
Dependent: difficult to replace without serious disruption.
Infrastructure: invisible until it fails, but essential to the system.

The companies with the strongest economics usually climb this ladder. They stop being a vendor and start becoming part of the operating environment.

Amazon Web Services is a classic example of this logic. A firm may not care deeply about a single software feature, but if its data, apps, and deployment pipelines live in the cloud, switching becomes expensive and risky. The same principle can apply to healthcare administration, payment rails, enterprise software, and now AI model access.

This framework also explains why “better” is not enough. A superior product that remains optional can lose to a merely good product that becomes infrastructural. In enterprise markets, integration beats admiration.

That is one reason why AI partnerships matter so much. They are not only about the model itself. They are about where the model lives, what chips train it, what cloud powers it, what enterprise tools expose it, and which workflows it quietly becomes part of.

If you want to know where value will accrue next, do not ask only, “What is the best technology?” Ask, “What is becoming unavoidable?”

The hidden tension: trust versus control

There is, however, an important tension in all of this. The more indispensable a system becomes, the more valuable it is, but also the more politically and operationally sensitive it becomes.

This is especially visible in AI and healthcare, two domains where trust is everything. People want systems that are powerful, but not opaque. Efficient, but not reckless. Automated, but not indifferent.

That is why the language of safety, reliability, compliance, and governance is not marketing fluff. It is the language of adoption. The moment a technology enters a workflow where mistakes cost real money or harm people, trust stops being a soft value and becomes an economic requirement.

In healthcare, trust determines whether members, providers, and regulators will accept a system’s role in managing risk and payment. In AI, trust determines whether enterprises allow a model into legal analysis, customer support, software development, or internal decision-making.

The paradox is that the same forces that create dominance also create scrutiny. The more a company becomes infrastructure, the more people ask whether it should be so powerful in the first place.

That is why the best strategic positions are often built around a careful balance: enough control to create differentiation, enough openness to create adoption, enough reliability to earn trust, and enough flexibility to survive regulation.

This balance is what separates durable platforms from overhyped bets. If a company cannot make outsiders comfortable depending on it, it will struggle to climb the dependency ladder.

Key Takeaways

Look for embeddedness, not just growth. The most powerful businesses are often the ones that become part of someone else’s essential workflow.
Trust is a moat. In high-stakes domains like healthcare and AI, safety, reliability, and auditability are not features, they are the gateway to adoption.
Distribution and infrastructure often matter more than product brilliance. A good product that is hard to replace can outperform a great product that sits outside the system.
Use the dependency ladder to evaluate companies. Ask whether a business is optional, integrated, dependent, or infrastructural.
Watch for middle-layer control. The biggest winners are often the firms that sit between fragmented users and fragmented providers, standardizing complexity for everyone else.

The real question is not who wins the race

It is easy to frame the future as a race for the best model, the best quarterly results, or the biggest headline. But those are surface metrics. The deeper question is more interesting: who becomes the layer the system cannot function without?

That question links healthcare, cloud computing, and AI into a single strategic map. The same pattern keeps repeating. Organizations that reduce uncertainty, coordinate complexity, and embed themselves in critical workflows do not merely compete. They become the environment.

And once a company becomes the environment, its power is no longer measured only by what it sells. It is measured by how much of the economy runs through it without thinking.

That is the real transformation underway. We are not just witnessing the rise of smarter tools or stronger balance sheets. We are watching the conversion of industries into infrastructure.

The firms that understand this will invest differently, partner differently, and build differently. Everyone else will keep mistaking product features for power, while the real winners quietly take control of the rails beneath them.