Out of Patients EP452:You Shouldn’t Need AI to Survive Cancer: Brad Power

The healthcare system keeps teaching cancer patients the same lesson. Learn fast or fall behind.

Brad Power understands that reality better than most people because he lived it. He entered cancer care as a patient with lymphoma and emerged from it with a realization that should concern every hospital executive, insurer, regulator, pharmaceutical company, and investor paying attention to where healthcare heads next. Patients increasingly trust artificial intelligence and distributed patient networks more than the institutions designed to guide them. That trend did not emerge from technological hype. It emerged from accumulated institutional failure.

Brad now runs Cancer Patient Lab and Open Cancer AI, projects built around a simple observation. Patients facing life threatening disease cannot wait for healthcare to become easier to navigate. They need information immediately. They need interpretation immediately. They need leverage immediately. Every delay carries biological and financial consequences. The modern healthcare industry keeps shifting administrative and cognitive labor onto patients while publicly framing that transfer as empowerment. Brad’s work sits directly inside that contradiction.

On paper, the AI revolution in healthcare sounds liberating. Give people better data. Give them AI copilots. Give them access to research once trapped behind institutional walls. Flatten information asymmetry. Democratize expertise. Then reality enters the room carrying a pathology report nobody can interpret, a prior authorization denial written in legal code, and a specialty pharmacy hold music playlist that sounds like someone trapped Phil Collins inside a fax machine.

Patients do not magically become sophisticated healthcare navigators because technology exists. They become terrified people trying to survive while decoding one of the most fragmented and economically distorted systems in American life. That distinction gets lost constantly in conversations about patient advocacy and digital empowerment. Healthcare organizations love telling patients to “be their own advocate” because the phrase sounds compassionate and modern. In practice, it often signals a system quietly outsourcing operational responsibility onto sick people and their families.

Banks still employ financial advisors because markets contain complexity. Airlines still employ pilots because aviation contains complexity. Yet cancer patients routinely hear some version of “you need to advocate for yourself” while receiving life altering diagnoses under severe emotional stress. Nobody hears that sentence during a functioning system. The phrase survives because the system financially rewards fragmentation while publicly marketing compassion.

Hospitals optimize throughput. Insurers optimize utilization management. Employers optimize premium exposure. Pharmaceutical companies optimize market share. Electronic medical record vendors optimize enterprise contracts. Every actor behaves rationally inside its incentive structure. Patients absorb the coordination burden left behind. Brad’s perspective becomes valuable because he refuses to romanticize that burden. He understands both the necessity and absurdity of forcing patients into operational command roles during catastrophic illness.

During our conversation, one tension kept surfacing beneath everything else. AI may soon provide cancer patients with more actionable clarity than large portions of the healthcare system itself. That statement sounds provocative until you examine the economics behind it. An AI model does not bill by CPT code. It does not operate inside a 12 minute visit window. It does not require prior authorization to summarize medical literature. It does not preserve referral patterns or protect opaque administrative structures that justify entire layers of middle management. AI processes information at scale. Healthcare institutions, by contrast, process incentives.

That distinction explains why patients increasingly upload scans, pathology findings, and lab reports into AI systems before hearing back from their care teams. Many clinicians find that behavior alarming. Some should. Hallucinations remain real risks. Oncology still requires human judgment, ethics, nuance, and accountability. Patients understand those risks better than institutions assume. What patients no longer tolerate is informational scarcity.

The old healthcare model relied heavily on asymmetry. Doctors knew things patients did not. Insurers understood reimbursement mechanics patients never saw. Hospital systems negotiated prices inside black boxes. Pharmaceutical economics remained almost entirely invisible outside industry walls. Digital access weakened that asymmetry. AI accelerates the collapse.

That collapse creates institutional discomfort because information control historically stabilized authority. Once patients independently cross reference treatment guidelines, compare survival data, understand formularies, or identify emerging clinical trials, they begin participating differently in their care. Sometimes they improve outcomes. Sometimes they create operational friction. Often both happen simultaneously. Healthcare systems still struggle to distinguish engaged patients from inconvenient patients.

That distinction carries enormous financial implications because informed patients often generate higher short term administrative costs. They ask more questions. They challenge denials. They request second opinions. They compare institutions. They scrutinize recommendations. They notice billing irregularities. They create work. Yet those same patients frequently avoid catastrophic downstream failures. They catch medication issues earlier. They escalate dangerous delays faster. They identify missing records. They reduce preventable utilization. They prevent expensive complications before those complications spiral into hospitalizations or worse.

American healthcare accounting rarely rewards those savings cleanly because the financial benefits scatter across disconnected entities while the administrative burden concentrates immediately. Fragmentation remains the business model. That reality explains why navigation technology now attracts serious investor attention. Whoever reduces friction inside healthcare coordination potentially unlocks enormous economic value. Administrative complexity consumes hundreds of billions of dollars annually depending on methodology. Every redundant authorization, delayed referral, missing record, duplicated scan, or abandoned follow up creates both financial leakage and human damage.

The danger comes when technology companies mistake information access for structural reform. Brad understands this tension well. Better tools help patients survive existing conditions. They do not automatically realign the conditions themselves.

Healthcare leaders often frame innovation around convenience. Faster scheduling. Better interfaces. Streamlined portals. AI chat support. Those improvements help, but they leave the underlying incentive architecture largely untouched. The system still rewards volume over continuity. It still rewards intervention over prevention. It still rewards opacity over transparency in many markets. It still places enormous cognitive labor on patients during periods of maximum distress.

Cancer patients experience those distortions acutely because oncology concentrates every institutional pressure point simultaneously. Insurance complexity, specialty pharmacy economics, hospital consolidation, clinical trial access, reimbursement incentives, workforce shortages, and financial toxicity collide directly inside cancer care. The result feels surreal to patients because survival increasingly depends on managerial competence unrelated to medicine itself.

A newly diagnosed patient now often needs some combination of scientific literacy, insurance literacy, digital literacy, financial literacy, negotiation skills, scheduling persistence, transportation logistics, employment protection knowledge, and psychological endurance simply to maintain continuity of care. Nobody trained for that role. Most people never expected to need those skills at all until the healthcare system quietly assigned them the job description.

Brad’s work acknowledges reality without pretending reality represents justice. That distinction matters because healthcare conversations often collapse into ideological theater that produces more outrage than structural insight. One side romanticizes markets. The other romanticizes centralized systems. Meanwhile patients sit on hold with specialty pharmacies trying to determine whether anti nausea medication requires another authorization form before chemotherapy starts Monday morning.

Real reform requires incentive realignment, not theatrical outrage. Employers already understand parts of this equation because they finance enormous portions of American healthcare indirectly through premiums and benefits administration. Every preventable delay, administrative burden, treatment disruption, or navigation failure eventually carries workforce and productivity consequences. The economic cost of patient confusion extends far beyond hospitals and insurers.

That broader economic reality creates an opening. Patient protection and economic efficiency increasingly align around the same target. Reduce friction. Reduce preventable failure. Reduce informational asymmetry. Reduce coordination burden. Reduce opaque administrative obstruction that generates downstream utilization and human damage.

Technology can support those goals. It cannot substitute for accountability. AI may eventually become a highly effective cognitive layer sitting between patients and institutional complexity. That possibility excites investors because scalable coordination creates scalable value. It should also unsettle healthcare incumbents because it exposes how much value currently derives from confusion itself.

Nobody willingly pays extra for friction once alternatives exist. That principle transformed banking, travel, retail, and media. Healthcare assumed regulatory insulation and institutional inertia would slow similar disruption indefinitely. Cancer patients do not have time for institutional pacing.

The deeper issue underlying Brad’s work involves legitimacy. Institutions maintain legitimacy when people believe expertise serves them competently and transparently. Legitimacy erodes when patients increasingly trust decentralized networks, online communities, and probabilistic AI tools more than formal systems carrying billion dollar infrastructure. Healthcare leaders should treat that shift as a warning signal, not a branding challenge.

Patients rarely abandon institutional trust casually. They abandon it after accumulated experiences teach them responsiveness, clarity, and coordination require personal escalation. I learned that lesson myself during brain cancer treatment at 21. My father handled the insurance fights because I physically could not. Most patients never see the hidden operational labor families perform behind catastrophic illness. They only experience the emotional exhaustion left behind.

Brad’s work exists because that exhaustion now intersects with powerful new technology. The opportunity ahead feels enormous. So does the risk. Used responsibly, AI could reduce cognitive overload, improve navigation, surface evidence faster, and help patients participate more effectively in care decisions. Used irresponsibly, it could create false certainty, widen disparities, commercialize vulnerability, and accelerate misinformation. The determining factor will not be the technology itself. The determining factor will be incentives.

If healthcare institutions continue protecting revenue structures built around opacity and fragmentation, patients will increasingly seek parallel systems outside traditional authority. If institutions instead embrace transparency, interoperability, navigation support, and patient aligned accountability, technology could strengthen trust rather than replace it.

Every industry eventually confronts the same question. Does the system reduce friction for itself or for the people depending on it? American healthcare answered that question decades ago. Patients finally noticed.