The Death of the Healthcare Disruptor

For nearly a decade, Silicon Valley has been promising to disrupt healthcare.

Billions of dollars have poured into artificial intelligence, virtual medicine, digital therapeutics, automated diagnostics, and technology-driven primary care. Investors were promised that healthcare would follow the same trajectory as banking, retail, transportation, and communications: software would replace expensive infrastructure, eliminate inefficiencies, and dramatically reduce costs.

It was a compelling story.

It was also, in many cases, fundamentally wrong.

Babylon Health, once valued in the billions, collapsed. Amazon shut down Amazon Care. Walmart closed its health centers. Forward, which raised hundreds of millions to reinvent primary care through technology, abruptly shut down. Pear Therapeutics, despite FDA-cleared digital treatments, filed for bankruptcy. Olive AI, once a multibillion-dollar healthcare automation company, sold off its core businesses.

These were not simply failures of execution. They exposed a deeper misunderstanding of how healthcare actually works.

Healthcare is not just a software problem. It is a delivery, distribution, trust, and incentive problem.

And artificial intelligence, paradoxically, may make traditional healthcare delivery systems more valuable—not less.

But there is a second, far more exciting possibility.

What if we stopped trying to replace healthcare with technology and instead used technology to give every person on Earth something most have never had?

A free, intelligent, lifelong health companion that knows their medical history, helps them understand their risks, monitors changes, and guides them toward better health—regardless of where they live or what they can afford.

And what if that same free platform could connect consumers with the pharmaceuticals, clinical trials, diagnostics, medical devices, and other products and services they actually need—creating an enormous economic opportunity without charging consumers or selling their personal health data?

That is the vision behind Global Health Atlas.

And it represents a fundamentally different way of thinking about the future of medicine.

The Billion-Dollar Graveyard

Consider what has happened to some of healthcare's most celebrated technology disruptors.

Babylon Health promised AI-powered healthcare at scale, combining symptom assessment, virtual consultations, and digital primary care. Its technology attracted enormous investor interest, but its business model struggled under the economic realities of delivering actual medical care. The company entered administration in 2023.

Amazon Care had something nearly every healthcare startup dreams of: virtually unlimited financial resources, sophisticated technology, and access to an enormous customer base. Yet Amazon discontinued the service in 2022, acknowledging that it was not a sufficiently complete offering for the enterprise customers it hoped to serve.

Walmart Health had another extraordinary advantage: physical locations, consumer trust, and proximity to millions of Americans. Yet Walmart closed its health centers in 2024, citing reimbursement challenges and rising operating costs.

Forward attempted to reinvent primary care with futuristic clinics, sophisticated diagnostics, and automated CarePods. It shut down in 2024.

Pear Therapeutics developed FDA-cleared prescription digital therapeutics, demonstrating that even clinically credible software can struggle when the reimbursement system has no straightforward way to pay for it.

Olive AI promised to automate healthcare administration, one of the industry's most expensive sources of inefficiency. Despite substantial funding and rapid growth, its original business ultimately unraveled.

These companies pursued different strategies, and their failures had different causes. But they shared a common challenge.

They underestimated the complexity of the healthcare delivery system.

Even the survivors illustrate the problem.

Teladoc generated approximately $2.5 billion in revenue in 2025 and facilitated millions of visits, yet continued to report substantial GAAP losses.

Scale alone did not turn healthcare into a high-margin software business.

The lesson is not that healthcare technology has failed.

The lesson is that technology alone is not healthcare.

Healthcare's Most Valuable Asset Is Not the Algorithm

Imagine an AI system identifies a potentially dangerous heart rhythm.

That identification may take milliseconds.

But what happens next?

Someone needs to determine whether the finding is clinically significant. Someone may need to order an electrocardiogram, arrange a cardiology consultation, review medications, obtain insurance authorization, coordinate treatment, and monitor the patient afterward.

The algorithm performed one task.

The healthcare system must perform the other ten.

This is what I call mechanism economics.

Economic value in healthcare is not confined to the intelligence that identifies a problem. It resides in the complete mechanism that turns knowledge into action.

That mechanism includes physicians, nurses, laboratories, imaging centers, hospitals, pharmacies, insurers, medical records, billing infrastructure, regulatory compliance, and, perhaps most importantly, trusted relationships.

A healthcare technology company can build an excellent algorithm in 18 months.

It cannot necessarily build a trusted network of 500 physicians, established payer contracts, decades of patient relationships, and a functioning referral system in the same period.

And here's the critical distinction:

A physician group can purchase world-class AI. An AI startup cannot simply purchase decades of clinical trust and integration.

This is why I believe the next generation of successful healthcare technology will be built around existing delivery systems rather than attempts to displace them.

AI will make physicians more productive, improve diagnostic support, reduce administrative burdens, and help manage patients between appointments.

But it will not eliminate the need for the physical and human infrastructure of medicine.

Nor should it.

When Intelligence Becomes Cheap, Everything Else Becomes More Valuable

There is another economic force at work, one that may be even more consequential.

I call it scarcity migration.

For decades, sophisticated medical knowledge was scarce.

Physicians spent years acquiring it. Hospitals organized around it. Patients paid to access it.

Today, AI is beginning to make sophisticated medical information and reasoning available at a fraction of their historical cost.

Foundation models from OpenAI, Anthropic, Google, and others are advancing at extraordinary speed.

As their capabilities improve, the cost of providing basic health information, interpreting medical data, and supporting clinical decisions should continue to decline.

This changes where value resides.

When intelligence is expensive, intelligence commands a premium.

When intelligence becomes inexpensive, value migrates toward what remains scarce.

In healthcare, that means trusted relationships, physical treatment, distribution, regulatory authority, continuity of care, and the ability to turn a recommendation into a real medical intervention.

A symptom-checking algorithm may eventually be a commodity.

A physician who knows your history, understands your circumstances, and can coordinate treatment will not be.

This is why the conventional Silicon Valley strategy of building a proprietary AI healthcare application and charging consumers a monthly subscription may prove increasingly difficult.

What happens when an equivalent AI model becomes available for pennies?

What happens when every insurer, hospital, medical group, and smartphone can offer similar intelligence?

The algorithm is no longer the competitive advantage.

The relationship and the infrastructure are.

But this same economic shift creates an extraordinary opportunity to deliver something that has historically been impossible.

Universal access to personalized health intelligence.

The Bigger Problem: Healthcare Was Designed to Treat Disease, Not Prevent It

We spend enormous sums treating diseases that, in many cases, could have been prevented, delayed, or better managed through earlier intervention.

Our healthcare systems are extraordinarily sophisticated at treating acute illness.

They are far less effective at helping individuals maintain health continuously over decades.

Consider the typical patient experience.

You see a physician once or twice a year, often for 15 or 20 minutes. Your physician reviews whatever information is available, addresses the immediate complaint, orders tests, prescribes medication, or refers you to a specialist.

Then you disappear from the system until the next appointment.

Meanwhile, your body is changing every day.

Your sleep patterns change. Your weight fluctuates. Your blood pressure rises or falls. Your glucose levels shift. Your medications interact. Your cardiovascular risk evolves.

Increasingly, wearable devices can capture some of these signals.

Yet most of this information never becomes part of a coherent, continuously updated picture of your health.

Your medical records may be scattered across several hospitals, specialists, laboratories, pharmacies, and insurance systems.

You may have access to a patient portal, but that does not mean you have a unified understanding of your health.

In many ways, the patient is the only person present for the entire healthcare journey—and often the person with the least access to the complete information.

The greatest opportunity in AI healthcare may not be replacing the doctor. It may be connecting the 8,760 hours in a year when the patient is not sitting in a doctor's office.

That is where prevention can become continuous rather than episodic.

And that is where artificial intelligence has the potential to change healthcare economics permanently.

Global Health Atlas: What If Everyone Had a Free AI Health Companion?

Imagine a world where every person has access to a free application that securely houses their health information.

Their medical history.

Their laboratory results.

Their prescriptions and medications.

Their vaccinations.

Their imaging reports.

Their family history.

Their wearable-device data.

Their personal health goals.

And, where available, their historical medical records across providers and healthcare systems.

Now imagine an AI companion that helps them understand that information, identifies meaningful changes, recognizes potential risks, explains medical findings in plain language, and encourages appropriate preventive action.

Not once a year.

Continuously, as new information becomes available.

This is the vision of Global Health Atlas.

Our objective is to develop a free, AI-powered personal health platform that can ultimately be made available to billions of people worldwide.

A platform where individuals—not hospitals, insurers, or technology companies—have meaningful control over their own health information.

A platform designed to help people understand their health before they become patients.

And a platform that works alongside existing healthcare providers rather than trying to replace them.

Consider a few examples.

A 55-year-old American uploads several years of bloodwork. The AI helps identify a worsening metabolic trend that deserves attention, explains why it matters, and suggests questions to discuss with a clinician.

A woman in rural India has limited access to primary care. Through a low-cost smartphone, she can access understandable health education, organize her medical history, and receive guidance on when professional evaluation is important.

A diabetic patient in Brazil connects a glucose monitor and medication history. The application helps track trends, identify possible concerns, and prepare useful information for the patient's medical team.

A patient in Kenya carries a portable health history that can be shared, with permission, when visiting a different clinic or hospital.

None of these examples requires AI to replace a physician.

All of them use AI to make healthcare more informed, more continuous, and more accessible.

Of course, medical AI must recognize uncertainty, protect sensitive data, and distinguish between education, decision support, and regulated medical care. A health companion should not present itself as an infallible diagnostician or substitute for emergency treatment.

The goal is not to automate medicine indiscriminately.

The goal is to democratize the information and guidance that help people make better health decisions.

Why Free Changes Everything

Most healthcare technology companies begin with a monetization question.

How much can we charge per patient?

What is our customer acquisition cost?

How do we maximize lifetime value?

How do we convert users into paid subscribers?

These are legitimate questions for commercial businesses.

But they can become obstacles when the objective is universal healthcare access.

Consider the billions of people living in regions with limited access to primary care, fragmented medical infrastructure, or inadequate insurance coverage.

A $20 monthly subscription might seem inexpensive in Silicon Valley.

For many households around the world, it is simply unaffordable.

Even $5 a month creates a barrier when the ambition is to reach billions of people.

Global Health Atlas begins with a different question:

What if the marginal cost of delivering personalized health intelligence becomes so low that charging the individual is no longer necessary?

AI inference costs have been declining dramatically, although secure storage, connectivity, clinical validation, support, and operations will continue to carry real costs.

If we can combine efficient AI models, cloud infrastructure, open technologies, philanthropic capital, and strategic distribution partnerships, we believe a free global platform becomes increasingly plausible.

The opportunity is not to build another healthcare subscription business.

It is to build shared health infrastructure.

Much as the internet made information broadly accessible and GPS transformed navigation, AI could help make personalized health guidance widely available.

The economics would be supported through philanthropic funding, technology partnerships, and sustainable commercial models that do not require selling personal health information or charging people simply to understand their own health.

Privacy and trust must be fundamental to the design.

Health information should be protected by strong security, meaningful consent, and clear limits on secondary use.

The user should be able to access, correct, export, and control the sharing of their information.

And a free service must never mean that the user's medical history becomes the product.

Free for Consumers. Enormous Economic Value for the Healthcare Industry.

There is another powerful economic opportunity in making Global Health Atlas free.

Free does not mean there is no commercial value. In fact, eliminating the consumer subscription may unlock a far larger economic opportunity.

Today, pharmaceutical companies, medical device manufacturers, clinical research organizations, diagnostics companies, and health and wellness brands spend billions trying to reach consumers who might benefit from their products and services.

Much of that spending is inefficient.

Companies advertise broadly, often with limited knowledge of whether the person seeing an advertisement has any medical need for the product being promoted.

Meanwhile, consumers are overwhelmed by healthcare marketing that may have little relevance to their actual health.

Imagine a different model.

A consumer using Global Health Atlas could voluntarily choose to receive information about products, treatments, services, or clinical research opportunities relevant to their individual health profile.

A person managing diabetes might discover a new glucose-monitoring technology.

Someone with a documented genetic condition might learn about an appropriate clinical trial.

A patient with cardiovascular risk factors might receive information about validated preventive programs, diagnostics, or therapies worth discussing with a physician.

A consumer interested in improving sleep might discover a relevant wearable device or evidence-based intervention.

For pharmaceutical companies, research institutions, medical device manufacturers, and healthcare providers, the ability to reach interested, appropriately matched consumers could be enormously valuable.

For consumers, it could mean discovering options they might otherwise never encounter.

The economic opportunity shifts from charging people to access their health information to helping them access the products, services, and innovations that may improve their health.

And the potential scale is extraordinary.

A free platform reaching hundreds of millions—or eventually billions—of people could become a powerful global channel connecting consumers with the healthcare ecosystem.

This could support several sustainable revenue models, including appropriately structured, consent-based referrals, clinical trial recruitment, relevant product discovery, and partnerships with healthcare organizations.

But there must be an inviolable principle:

The consumer owns and controls their health information. Their identifiable medical data is never sold to advertisers, pharmaceutical companies, or other commercial partners.

Commercial participation should be opt-in, transparent, privacy-preserving, and subject to appropriate clinical and regulatory safeguards. Companies should not be able to purchase favorable medical recommendations, and clinical trial opportunities must follow applicable consent and research protections.

AI-generated health guidance must remain independent of commercial incentives.

This is how Global Health Atlas can align incentives rather than repeat the mistakes of traditional healthcare.

Consumers receive a free, intelligent health companion.

Physicians receive better-informed patients.

Researchers gain a potential pathway to reach consenting, eligible trial participants.

Healthcare innovators can connect with people who may genuinely benefit from their products.

And commercial participation can help support the infrastructure required to keep the platform free.

The larger the network becomes, the more useful it can be to consumers—and the more valuable it can become to the healthcare ecosystem.

The goal is not to monetize illness.

It is to create an economic model in which helping people identify and address their health needs supports universal access to health intelligence.

That may ultimately be a more scalable and sustainable model than charging every patient a monthly subscription.

The Most Powerful Distribution Network May Already Exist

One of the most expensive mistakes in healthcare technology has been attempting to build distribution from scratch.

Companies spend enormous amounts on advertising, sales teams, employer contracts, and customer acquisition.

But billions of people already have access to smartphones.

Global technology companies already operate communication networks and digital platforms with extraordinary reach.

Healthcare organizations already have relationships with patients.

Public-health systems, nonprofits, and local community organizations already serve populations that commercial healthcare startups often overlook.

Why recreate these networks?

Global Health Atlas is exploring a partnership model that could bring together AI developers, cloud infrastructure providers, satellite connectivity companies, device manufacturers, medical institutions, governments, and philanthropic organizations.

Imagine the potential reach of distribution through existing global platforms.

Imagine connecting personal health records with data from Garmin, Apple, Oura, WHOOP, and other devices where users choose to participate.

Imagine offering a useful health companion in multiple languages, including regions where physician access is limited.

And imagine doing it without requiring each person to purchase another healthcare subscription.

The opportunity is enormous.

But it requires a different mindset.

Instead of acquiring customers, we need to reach people. Instead of maximizing revenue per patient, we need to maximize useful health outcomes per dollar spent.

That is a very different objective function from the one that drives most venture-backed healthcare startups.

A 100-Country Expedition to Test the Idea

Technology built in Silicon Valley does not automatically work in rural Africa, Southeast Asia, Central Asia, or remote island communities.

Healthcare access varies enormously.

So do languages, infrastructure, medical practices, cultural expectations, connectivity, and trust.

A global health platform cannot simply be designed in a conference room and declared universal.

It needs to be tested in the real world.

That is why Global Health Atlas is planning an approximately 100-country expedition beginning in February 2028.

The mission will travel across dramatically different healthcare environments to engage with clinicians, public-health organizations, community leaders, technology partners, and individuals.

Our objective is to help launch, test, document, and refine the platform while learning what actually works.

Where does AI provide meaningful value?

Where does it fail?

How can a platform serve people with intermittent connectivity?

How do we make personal health records useful when historical medical data is unavailable?

How do we adapt to local languages and health systems?

How do we build trust among populations that have never used digital health services?

And how do we ensure that technology improves access without creating new forms of exclusion?

These questions cannot all be answered by engineers.

They require clinicians, researchers, public-health professionals, community organizations, and the people who will ultimately use the platform.

The expedition is intended to be part of that learning process.

The Real Revolution Is Not Replacing Doctors

The healthcare industry's fixation on disruption has distracted us from a much larger opportunity.

We do not need another company promising to eliminate physicians.

We need technology that makes physicians more effective.

We do not need another proprietary medical algorithm hidden behind a subscription.

We need affordable, trustworthy intelligence that helps people understand their own health.

We do not need more fragmented applications that each capture one small piece of a patient's medical history.

We need an integrated, lifelong picture of health that individuals can carry with them.

And we do not need to force the world's poorest populations into a commercial healthcare model designed around affluent consumers.

We need a system built around access, prevention, and human well-being.

There will always be a critical role for hospitals, medical groups, specialists, nurses, and other healthcare professionals.

Indeed, as AI becomes more capable, the importance of those institutions may increase because they possess what algorithms cannot independently provide: accountable clinical judgment, physical intervention, and human relationships.

But outside the walls of those institutions lies a vast, largely unserved opportunity.

Billions of people who need better health information.

Millions who struggle to navigate fragmented systems.

Countless preventable complications that might be recognized earlier with better monitoring, education, and follow-through.

That is where the next revolution belongs.

From Healthcare Disruption to Healthcare Empowerment

The first generation of healthcare disruptors tried to use technology to capture the economics of healthcare.

The next generation has an opportunity to use technology to fundamentally improve those economics.

The difference is profound.

One model asks how much money can be extracted from delivering care.

The other asks how much suffering, disease, and unnecessary expense can be avoided by making health intelligence broadly accessible.

AI gives us an unprecedented opportunity to pursue the second model.

Not because artificial intelligence can do everything a physician does.

But because it may soon be inexpensive enough to provide billions of people with information, monitoring, and decision support that were previously unavailable to them.

And for the first time, we have the opportunity to build a global healthcare ecosystem where consumers receive personalized health intelligence for free, healthcare companies profit by delivering genuinely relevant solutions, and economic incentives support prevention rather than simply the treatment of disease.

That is a much bigger opportunity than another virtual-care company.

And potentially a much more important one.

The next great healthcare innovation may not be a billion-dollar company.

It may be a free application that helps billions of people live longer, healthier lives.

That is the future we are working toward at Global Health Atlas.

And it is a future worth building.