I. The ghost on the deck

There is a story I have told, in one form or another, for a long time. It is a story about a ship.

In 1747, aboard HMS Salisbury, a young naval surgeon named James Lind took twelve sailors whose gums had blackened and whose limbs had begun to fail them, and he divided them into pairs. To one pair he gave cider. To another, dilute sulfuric acid. To others, vinegar, seawater, a purgative paste of tamarind and cream of tartar. To the last pair he gave two oranges and a lemon a day. Within six days, the citrus pair was nearly recovered. The others were not. Lind had, almost without meaning to, invented the controlled trial. He had also, in the same gesture, conjured a ghost that would haunt medicine for the next three centuries: the Average Patient. A statistical phantom that exists nowhere except in spreadsheets, and yet around whom we built every regulatory fortress, every approval pathway, every billion-dollar trial that followed.

I told this story once, years ago, to a young contractor at the FDA at a meeting about something else entirely. His name was Jeremy Walsh. He listened the way he always does, carefully and without performance, and when I had finished, he asked the question I have come to recognize as his particular gift. He asked it again today on a stage at White Oak, MD, as the FDA’s first Chief AI Officer, in front of a room full of cameras, reporters, and pharmaceutical executives:

Wait. We are still doing it that way?!

In our meeting years back, I told him that yes, more or less, we were. That what changed between the Salisbury and the modern clinical trial ecosystem was not the basic shape of the enterprise but the scale at which we did it and the bureaucratic stratigraphy we had laid down on top of it. Sponsors run trials. Sites collect data. The data is packaged, cleaned, narrativized, and shipped across in pages of PDF, thousands of pages in a typical IND, some of it not even human-readable. Counts of counts of counts. Tables of tables of tables. The regulator sees the data only after the sponsor has finished telling itself the story of what the data means. Lind would have recognized the structure. He would not have recognized the latency. In a way, we have actually regressed.

My conversation with Jermey became a white paper. The white paper went up the chain. Nothing happened, which is what usually happens, and Jeremy went back to his other work. But he kept asking the question. He asked it for years. He asked it as a contractor and he asked it after he became Chief AI Officer at the FDA and he asked it in every meeting where someone told him the way things were done was the way they had to be done. Today he got to answer it.

II. What was announced

Today, the FDA announced the real-time clinical trial initiative involving AstraZeneca’s TRAVERSE study in mantle cell lymphoma, with sites at MD Anderson and Penn and Amgen’s STREAM-SCLC trial in small cell lung cancer. A Request for Information also opened today. A broader pilot program launches this summer. Commissioner Marty Makary, who led the announcement, framed the rationale precisely. For sixty years, he said, we have been conducting clinical trials in essentially the same way, and the lag between data and decision has not been a feature of safety. It has been sediment. The accumulated weight of a system that was built when data lived in filing cabinets and submissions arrived by truck. The agency, he said, can reduce that lag without cutting any corners on safety, which remains the FDA’s number one priority. He said the line plainly, and he said it more than once.

This is the part of the announcement that most deserves attention, because it is the part that is easiest to misread. What the FDA is doing is not deregulation. It is a re-engineering. The protocols are still protocols. The investigators still consent patients. The reviewers still review. Paul Burton, Chief Medical Officer at Amgen, was deliberate on this point. Pragmatic real-time clinical trials, he said, do not mean clinical research with less rigor. The fundamentals of good science do not change. Protocol development, protocol oversight, informed consent, monitoring, strong governance, FDA review. All of it remains absolutely standard and absolutely essential. Patient safety and data integrity, he said, will always be the north star.

That language matters because the architecture being introduced is one in which safety is not preserved by latency but expressed through signal. Today’s announcement is, in this sense, a strengthening of efficacy-signal detection and pharmacovigilance, not a relaxation of them. Makary made the point directly in the press Q&A. Safety signals and clinical endpoints are agreed upon by the sponsor and the FDA in advance, as they have always been. What changes is that those signals can now be transmitted to the agency in real time rather than packaged into a submission months or years later. During the pilot, sponsors will continue to submit data through the conventional pathway in parallel, so that the agency can compare the two channels and refine the model. Nothing has been removed from the safety apparatus. Something has been added.

What changes, then, is the latency. What changes is the assumption, baked so deeply into the architecture of clinical development that we forgot it was an assumption, that the regulator should see the data only after the sponsor has packaged it. Jeremy described the new model on stage in language that I suspect will sound radical to people outside the field and obvious to people inside it. We do not need everything we have ever seen. We need signals. We need to define, with the sponsor and the FDA together, what would actually change a regulatory decision, and we need to see those things happen as they happen. Not a year later. Not at the next data lock. Now.

III. Listening to the patient instead of to the form

The first time I worked with Jeremy was at the National Cancer Institute, where we designed a clinical trial in adult patients with cancer using voice recognition and facial recognition to estimate pain and mood. The patient sat down. The cameras and microphones did the rest. No survey. No clinician translating an expression into a number on a Likert scale. The signal came directly from the human being in the room, in real time, captured continuously, with a fidelity that no questionnaire has ever achieved or could. That dataset, to my knowledge, remains the largest of its kind.

The deeper observation that emerged from that work, and that I think is the actual intellectual core of what was announced today, is that the data we collect in clinical trials is not always a faithful image of the patient. It is a heavily compressed, heavily edited, heavily delayed reconstruction. The patient experiences the disease in continuous time. We sample it in discrete time. The patient lives in signal. We submit in summary. Every step in the pipeline from bedside to FDA reviewer is a lossy compression, and the loss is not random. It systematically privileges what is easy to measure and structure over what is meaningful and continuous. The Likert scale erases the wince. The case report form erases the silence before the patient answers. The quarterly data lock erases the morning the toxicity actually appeared.

What changes when signals flow in real time is not just the speed of regulatory decisions, although that matters and it matters enormously to the patient who is in hospice waiting for the drug. Jennifer Litton, the Chief Clinical Research Officer at MD Anderson and one of the academic leaders behind the AstraZeneca trial, made the point with the kind of moral clarity that only someone who has stood at the bedside can make. She is an oncologist. She is the daughter of two parents with cancer. She was diagnosed with cancer this year, and finished surgery and radiation. When you do not know whether the next patient who needs a clinical trial is going to be you or someone you love, she said, you want that trial available right now. The latency in our current system is not abstract. It is measured in funerals.

What changes, then, is also the resolution of what the regulator can see. We move from low-frame-rate snapshots of an aggregate to a continuous record of individuals. The Average Patient, the ghost we summoned on the deck of the Salisbury because we had no other way to reason about populations at scale, becomes optional. Not abolished. Optional. We can still aggregate when aggregation is what the question requires. But we are no longer forced to aggregate as a precondition of being able to see anything at all.

IV. The renovation

There is a temptation, when an announcement of this kind lands, to declare that the old order has fallen and a new one has taken its place. That is not what has happened today. Two trials. One technology partner (Paradigm Health). This is the first sentence of a new chapter, not the chapter itself. The hard work, the transformation, is the scale-up. It is the second sponsor, and the tenth, and the hundredth. It is the standards work that has to happen quietly behind the announcements, the EHR-to-EDC plumbing, the sponsor data systems that have to be re-architected, the reviewer workflows that have to be redesigned around streams instead of submissions. It is, as one industry veteran told me afterward, the long stretch where we will have to run both systems in parallel, the old way and the new way at once, before the old way can finally be retired.

There is also the human dimension, which is easy to miss in the technical conversation. Emma Meagher, who runs clinical and translational research at Penn and is one of the principal investigators on the AstraZeneca trial, made an interesting observation. She has been a trialist on the front lines for thirty-two years. She knows, better than almost anyone, how punishing the current process is on the people who actually run it. The research coordinators. The trial nurses. The project managers. They are virtually impossible to find and harder to keep. What this innovation does, she said, is not just shorten the FDA’s timelines. It eases the academic interface of patient care. It gives the workforce that sustains clinical research a reason to stay. The renovation is not only of the regulatory architecture. It is of the human one.

What Makary said almost in passing was, I think, the line that we should remember. He told the room, with a self-deprecating shrug, that he was not going to show a PowerPoint, because PowerPoint had traumatized him over the course of his career. He showed one graphic. The drug development and approval process, beginning to end. Each phase a box. Each box, he said, currently being re-engineered. The graphic is the same graphic we have been showing for sixty years. The boxes are the same boxes. What changes now is what flows between them, and how fast, and at what fidelity, and to whom.

This is not a revolution. Revolutions tear down. This is something rarer and harder. This is a renovation of a working building, conducted while the building remains occupied by patients who are, right now, today, waiting for the next box to open. The ghost of the Salisbury, the Average Patient who never existed, can finally begin to be retired. Not because we have stopped caring about populations, but because we have learned, at long last, how to listen to individuals in real time, and to aggregate that listening without losing it.

To Jeremy, congratulations. He kept asking the question for years after most people would have stopped. To the team at FDA, the sponsors, the academic centers, the technology partners who made this real, congratulations. To everyone watching from outside the agency wondering whether this is a small thing or a large thing, I will say this. It is an important thing. And in ten years it can hopefully look like the thing that broke the dam.

The signal is moving. The ship has finally left the harbor. Now we have to learn to navigate.

Subscribe now

Every oncology drug approved based on tumor-based endpoints in the last twenty-five years rests on an instrument. The instrument is not a machine, not a molecule, not a laboratory assay. It is a rather arbitrary convention. A rule about how to press a ruler against a shadow on a screen and decide whether the shadow has grown. The convention is called RECIST (Response Evaluation Criteria in Solid Tumors) and it dictates, with the stubborn authority of orthodoxy, how major regulatory decisions in solid tumor oncology have been made since the late 1990s.

RECIST works like this: A radiologist examines a CT scan and selects up to five “target lesions,” the largest measurable tumors in a given organ. She measures each one along its longest diameter, sums those diameters, and compares the sum to a baseline. If the sum has shrunk by 30% or more, she writes partial response. If it has grown by 20% or more, she writes progressive disease. Anything in between is stable disease. Complete disappearance is a complete response. These four words, expanded into the binary of ORR (objective response rate: the fraction of patients who achieve complete or partial response), are what we submit to the FDA as evidence of efficacy. They are the basis on which drugs are approved, priced, and prescribed to patients with cancer.

The framework has the unusual property of being both ubiquitous and invisible. Few oncologists have heard of it. Almost no one in clinical practice actually uses it. It is the instrument by which drugs are certified, not the instrument by which they are administered. This is the first fact about RECIST worth holding onto. We will return to it.

I. Two readers, one scan

In 2017, while at the FDA, my colleagues and I pooled the paired radiology assessments from a decade of solid tumor registration trials. The structure of these trials has a quirk useful to our purpose: for regulatory submission, every scan is read independently by two qualified radiologists, one primary investigator review and the other a blinded radiologist at a central reading facility, each blinded to the other’s work. If they disagree, an adjudicator breaks the tie. The raw data, kept for internal auditing, records how often two expert readers, looking at identical images and applying identical criteria, reach identical conclusions.

Our pooled sample contained 13,677 paired assessments. The raw concordance rate was 69.5 percent. Two of every three paired reads agreed. The remaining third did not. In hard-to-measure malignancies such as pancreatic and ovarian cancers, the discordance exceeded 40% and, in some cases, reached the predictability of a coin flip.

This figure is higher than it ought to be. In the language of measurement theory, it corresponds to a weighted kappa somewhere between 0.30 and 0.55. Kappa is a statistic that asks how much two raters agree beyond what you would expect from chance alone; a value of 1.0 is perfect agreement, 0.0 is no better than coin flips. The number we observed sits at the lower boundary of what Landis and Koch, in their widely cited 1977 classification, called “moderate.” Moderate is what you would expect from a diagnostic tool in early development. It is not what you would expect from the instrument on which the entire oncology drug development enterprise rests.

The standard defense of the framework, when this figure is raised, is procedural. Discordance, the argument goes, is non-differential. It is distributed symmetrically across treatment and control arms. Therefore it does not invalidate the comparison. The p-value remains honest. This is partly true for randomized studies. It is also profoundly incomplete.

II. The attenuation

Here is what the procedural defense leaves out, beyond the fact that many cancer drugs are advanced in single-arm studies. When you inject random error into a comparison of two proportions, the observed difference between the proportions does not stay the same. It shrinks. This is a known result in epidemiology, tracing at least to the work of Irwin Bross in the 1950s, but it is rarely invoked in discussions of clinical trial design because most people prefer to believe that their measurements are not noisy.

The mathematics is merciful in its clarity. If the true response rate in the treatment arm is P(T) and the true response rate in the control arm is P(C), and each patient is misclassified with probability m, the observed difference between arms is the true difference multiplied by the factor (1 - 2m). At our measured discordance rate of thirty percent, that factor is 0.40.

A drug whose true advantage over standard of care is twenty percentage points of ORR, a substantial effect by any reasonable standard, appears in the trial record as a drug with an eight-point advantage. The biological signal has been attenuated by sixty percent before the statistician ever sees it.

Attenuation increases the sample size. Because the number of patients required to detect a given effect scales as the inverse square of that effect, a trial that would need two hundred patients to detect the true difference would need, under this measurement regime, approximately one thousand two hundred and fifty to detect the attenuated one. Most trials do not enroll 1,250 patients. Most trials are powered to detect the signal they were designed to detect, assuming the measurement faithfully transmits it.

The directional consequence is the one that matters, and it is rarely stated plainly. Random measurement error biases observed effects toward the null hypothesis. At the level of the measurement substrate, this produces a systematic excess of false negatives over false positives. A trial that fails to reach statistical significance is not necessarily a trial of an inactive drug. Some meaningful fraction of negative trials in solid tumor oncology are trials of active drugs evaluated with an instrument that could not see them.

No one knows the size of that fraction. The instrument, by its nature, erases the evidence of what it failed to detect.

III. Four boxes for a continuous world

The second property of RECIST is a property of encoding rather than of measurement. The underlying biological variable, the percent change in tumor burden over time, is continuous. It takes any value on the real line, bounded below by minus one hundred percent and above by whatever the tumor can manage. The framework transforms this continuous variable into an ordinal category with four levels, and then, for ORR, collapses those four levels into a binary. Something biologically rich becomes something statistically thin.

There is a way to quantify the loss, borrowed from information theory. A continuous percent-change measurement carries, under plausible distributional assumptions, between 5 and 8 bits of information per patient per scan. (A bit is the unit Claude Shannon introduced in 1948 to measure the content of a signal: one bit is the information gained from the outcome of a fair coin flip.) A four-level ordinal carries two bits, which is the base-2 logarithm of 4. A binary outcome carries one. The RECIST encoding, at its most generous, transmits at most a quarter of the information contained in the measurement it encodes. At ORR, it transmits less.

The effect is not merely a loss of resolution but a loss of biological distinguishability. An agent that produces a uniform 25% shrinkage in every treated patient, a remarkable result implying a real mechanism of action operating reliably across a population, can yield an ORR of zero. It is categorized, on the endpoint, as indistinguishable from placebo/no treatment. It is also indistinguishable, at the endpoint, from an agent that produces 19% growth. Three biologically distinct populations, trivially separable in the underlying distribution of percent change, collapse to the same point in the reported outcome.

This is the structural consequence of mapping a continuous variable through a step function. The step does what steps do. It discards what lies between its treads.

IV. Stochastic at the edge

The thresholds themselves introduce a third pathology, one that becomes vivid if you imagine the patient whose true response is not in the middle of a category but at its boundary.

For example, let’s look at a patient whose tumor burden has decreased by 28%. Two percentage points below the 30% partial response threshold. Stable disease by the rules, but barely. Suppose the inter-reader standard deviation on the sum of diameters, consistent with the discordance data, is about ten percentage points. The probability that a given reader assigns this patient to the PR category, under a normal approximation, is the probability that the reader’s estimate falls below -30%, which works out to about 42%.

Forty-two percent. The same biology, the same scan, the same criteria, yields partial response in forty-two of one hundred reads and stable disease in fifty-eight. The categorical outcome is, at the boundary, approximately a coin flip. And the reported trial effect depends on how many of these coin flips land on the side that counts.

This threshold instability does more than add variance. It creates a selection pressure that propagates backward into the development pipeline. Because ORR binarizes sharply at minus thirty percent, a sponsor who wishes to maximize the probability of clearing that threshold is rewarded for advancing agents capable of producing deep responses in a subset of patients. Agents that produce shallow but consistent cytoreduction across a population, even when the underlying biology is meaningful, are structurally disfavored. They cannot clear the threshold often enough to generate an ORR that the machinery can see. And our machinery is impatient. When meaningful tumor responses are not seen in early clinical trials, the entire program is at risk.

The shape of the modern oncology pipeline, its preferential enrichment for drugs that drive responses visible to the human eye, is a direct consequence of our endpoints. We have been selecting not for what can prolong survival or increase quality of life, but for what our instruments can detect.

V. The instrument that does not exist outside the trial

The third property is the most easily stated and the most commonly overlooked. RECIST is a rather artificial clinical trial instrument. It is not a clinical tool. Community oncologists, the people who actually treat the overwhelming majority of patients with cancer, do not compute sums of diameters. They do not segment target lesions into protocol-defined categories. They read narrative radiology reports written in ordinary English (”stable appearance of the right upper lobe mass”) and make decisions on the basis of clinical judgment integrated with that narrative.

Which is to say: the endpoint that certifies a drug’s efficacy in the trial setting is not the endpoint that governs its use in the practice setting. The variable on which approval rests is not present in the data that routine care generates. This is the oncology equivalent of approving, say, statins on the basis of LDL reduction only to find out that community physicians use blood pressure to administer and titrate therapy.

When we attempt to validate oncology clinical trial findings against real-world evidence, using electronic health records and community radiology reports, we cannot reconstruct a RECIST assessment. The comparison has to be made at a different level of abstraction, using different variables, and the correspondence between the two settings, at the level of the individual patient or the individual drug, remains structurally unverifiable.

The conventional response to this observation is that RECIST is a surrogate for clinical benefit rather than the benefit itself. A surrogate is a measurable variable used as a proxy for the outcome of real interest, which in oncology is almost always overall survival. The real question, on this view, is whether RECIST-based endpoints predict overall survival. The meta-analytic answer, accumulated across hundreds of trials and now reasonably mature, is that they do so imperfectly. PFS-to-OS correlations in solid tumor oncology routinely yield coefficients of determination below 0.5. ORR-to-OS correlations are lower. The surrogate explains less than half of the variance in the outcome it is supposed to predict.

The framework persists, then, less because of its predictive validity than because of its institutional entrenchment. It is the thing we all agreed to use. Agreement has its own gravitational pull, and for a long time the conditions that would permit disagreement did not exist.

VI. What the endpoint could not see

An instrument with 30% reader discordance, indicating moderate agreement by standard statistical classification, can attenuate observed effects by 60% and inflate required sample sizes sixfold. The attenuated signal is then projected onto an ordinal encoding that transmits at most two bits per scan and produces stochastic categorization at the thresholds where the reported effect is determined. The resulting endpoint is computable only within the trial setting and cannot be reconstructed from the data generated by clinical practice after approval.

The directional bias of the whole apparatus is toward false negatives. It favors drugs whose mechanism produces a small number of responses that are deep enough to be detected by the human eye over drugs whose mechanism produces a large number of modest responses, but can prolong survival or improve quality of life. It cannot see agents whose signals fall below the threshold for categorical instability, and it cannot distinguish, at the endpoint, between agents that differ by biologically meaningful amounts.

The statistical machinery layered on top of this substrate, all the careful p-value thresholds and hierarchical testing procedures and alpha-spending rules, is bounded by what the substrate permits. No refinement of the hypothesis test can recover information that the measurement did not transmit.

The cost is not just borne by the sponsors whose trials failed. It is borne by the patients whose therapies were never developed because the endpoint could not see them.

VII. The substrate is replaceable now

The conditions that produced RECIST in the late 1990s are no longer operative. Volumetric segmentation of tumors, computationally prohibitive a few years ago, is now a solved problem. Foundation models for medical imaging, trained on orders of magnitude more data than any individual radiologist will read in a career, are mature. Automated assessment at scale, across both trial and community settings, is a deployed capability.

Over the past several years, I have been working on a replacement substrate at Project Data Sphere. We call it the Total Tumor Burden Index. It is volumetric rather than diametric, which reduces reader-dependent variance. It is continuous rather than categorical, which eliminates both the information loss of ordinal encoding and the instability at categorical boundaries. It is machine-generated rather than reader-dependent, which removes the dominant contemporary source of measurement noise and allows deterministic reproducibility across sites and across re-reads. And it is equally deployable in trial and community settings, which means the trial measurement and the practice measurement can, for the first time, be the same variable.

The regulatory path will take time. Prospective comparison with RECIST in contemporary datasets will take time. Integration across imaging infrastructure will take time. None of these barriers is technical. The technical conditions for the replacement are being satisfied with increasing precision.

VIII. Coda

The measurement substrate in solid-tumor drug development is not neutral. It has been shaping the pipeline for twenty-five years in ways the field has not made fully explicit to itself. The attention has remained on the superstructure, the p-values, the statistical analysis plans, and the regulatory pathways, when the leverage is actually in the layer underneath. Replacing the measurement, rather than refining the statistics computed upon it, is the tractable path forward. The conditions for that replacement are, at last, in place.

What we measure determines what we can find. What we can find determines what we develop. What we develop determines who is treated. The ruler we have been pressing against the tumor for a generation is not the only ruler available. It is simply the one we inherited. And that inheritance is due for review.

Share

There is a principle in pharmacology that took medicine decades to fully absorb. Liposomal doxorubicin and free doxorubicin are the same molecule. Yet one accumulates preferentially in tumors while the other damages the heart. The liposome does not make the drug more potent. It changes where the drug goes, what it encounters, what it spares; by exploiting the leaky vasculature of the tumor microenvironment and staying encapsulated in the tighter junctions of cardiac tissue. The container shapes what the active compound can do.

The same principle, it turns out, applies to AI systems. And a paper published at the end of March 2026 by researchers at Stanford, MIT, and KRAFTON gives us the clearest demonstration yet of what that means in practice and what it makes possible.

What the paper found

The paper is called Meta-Harness: End-to-End Optimization of Model Harnesses. Its subject is the code that wraps a large language model, the layer that determines what information the model sees, in what order, retrieved from where, and framed how. The authors call this the harness. It encompasses three things: memory, meaning what the system retains between interactions; retrieval, meaning what it fetches from external sources at each decision point; and prompting, meaning how context is assembled and sequenced before the model reasons about it.

The central observation is straightforward. Research cited by the authors has shown that changing only the harness around a frozen, untouched model can shift performance by a factor of six on the same benchmark. Same weights. Different wrapper. Until now, harnesses have been built entirely by hand: engineers designing prompting frameworks, retrieval strategies, and memory structures through iteration and intuition, exploring a small fraction of what is possible.

Meta-Harness automates this process. It works through a recursive loop. A candidate harness is proposed, evaluated against a set of tasks, and everything is saved in raw, uncompressed form, not a score, not a summary, but the full transcript of every prompt the model received, every response it gave, every tool call it made. A coding agent then navigates that filesystem, reads what it needs, and proposes an improved harness. The loop repeats roughly sixty times in a full run.

What distinguishes this from every prior approach to automated optimization is the refusal to compress. Systems like OPRO, TextGrad, and AlphaEvolve hand the optimizer a digest: a scalar score, a short critique, a recent snippet. Meta-Harness gives the agent the raw logs and trusts it to navigate selectively, reading a median of 82 files per iteration across the full history of prior attempts. That access to raw diagnostic history, rather than compressed summaries, turns out to be the key design decision.

The results across three domains are real and substantial. On a text classification task where a model updates its memory from labeled examples and is evaluated on held-out cases, the discovered harness outperforms the best hand-designed system by 7.7 accuracy points while using four times fewer tokens. On 200 olympiad-level mathematics problems, a single harness discovered during search improves accuracy by an average of 4.7 points across five AI models never seen during the search process, indicating the discovered strategy is genuinely general. On TerminalBench-2, a benchmark of 89 real-world tasks requiring AI agents to autonomously compile code, debug systems, and complete multi-step terminal workflows with full automated verification, the discovered harness places first among all agents running on the same underlying model, surpassing every hand-engineered competitor.

But what the paper demonstrates is not just that automated harness search works. It is that the agent doing the searching behaves in ways that matter for how we think about these systems going forward.

The agent that reads its own case files

The paper’s appendix traces, iteration by iteration, how the agent reasoned through a live search run on TerminalBench-2, starting from a strong hand-engineered baseline. What emerges is not a description of brute-force search. It is a description of systematic diagnosis.

In the first two iterations, the agent bundled a structural bugfix with a rewritten prompt template. Both regressed from the baseline. In the third iteration, rather than proposing a third fix, the agent looked back across both failures. It noticed that while the two attempts differed in their structural changes, they shared one intervention: the prompt rewrite. It separated the confounded variables and wrote verbatim into its log: “the structural bugfixes were confounded with harmful prompt changes.”

This is abductive reasoning, the same logic a careful physician applies when two patients on similar regimens both deteriorate. You do not conclude the drug failed. You ask what they had in common. You isolate. You test. The agent was performing differential diagnosis on its own previous attempts, using the filesystem as its case history and the execution traces as its clinical notes. It could only do this because it had access to the raw logs. A compressed summary would have hidden the shared intervention entirely.

By the seventh iteration, after six consecutive regressions that taught the agent where the system’s fragility lived, it stopped modifying the existing machinery and instead added something before the loop began: a snapshot of the computational environment injected into the initial prompt. Its recorded reasoning: avoid touching the previously fragile components. It had learned not just what failed but why and used that understanding to reason about risk.

This behavior (reading raw history, identifying confounds, forming causal hypotheses, reasoning about where fragility lives) is what full access to prior execution traces enables. Compressed feedback cannot support it. And it is precisely this kind of reasoning that the hand-engineering of clinical AI harnesses has always depended on human experts to provide.

The compression problem, demonstrated

To understand why this matters for biomedicine specifically, it helps to look at what happens when LLM harnesses are hand-engineered in a clinical setting.

In a study we conducted last year at Project Data Sphere with Thomas Jefferson University, four independent investigators used LLMs (ChatGPT-4 and Gemini Advanced) to analyze data from a clinical trial evaluating LY2510924 combined with carboplatin and etoposide for extensive-stage small cell lung cancer. All four investigators worked from the same dataset, the same chain-of-thought prompting framework, and the same analytical objectives.

The results were interesting but inconsistent. One investigator found a statistically significant improvement in progression-free survival with the addition of LY2510924, contradicting the original publication and the other three investigators. Another reported an objective response rate that deviated substantially from every other analysis. Cohort sizes ranged from 89 to 180 participants from the same dataset, suggesting the LLMs characterized entirely different patient populations depending on how the prompts were constructed. The conclusion of the paper was clear: variability in prompting strategies drove variability in clinical conclusions, and highly trained subject matter experts remain essential to review and interpret LLM-generated analyses.

This is the harness problem in clinical research, demonstrated under controlled conditions. The chain-of-thought framework was the harness. It was hand-engineered. Different investigators interpreted it differently. The model was sensitive to subtle prompt variations. And the outputs, including whether a drug appeared to improve survival, changed accordingly.

Meta-Harness is a direct response to exactly this problem. Rather than asking human experts to design the optimal prompting framework and then hoping investigators apply it consistently, automated harness search would systematically explore the space of prompting and retrieval strategies, read its own failures at the case level, this analysis got PFS wrong, what was it shown, what did it retrieve, what framing led it astray, and discover configurations that produce consistent, reproducible outputs across investigators, model versions, and datasets. The subject matter expert moves from designing the wrapper to evaluating and validating what the automated search discovers. That is a meaningful and practical shift in how we deploy LLM-based clinical tools.

What the 40% tells us

A drug achieves a 40% ORR in a clinical trial. We call this a success. The drug is approved, deployed, and prescribed. But 60% of patients who receive it do not respond. They experience toxicity without benefit, akin to taking a toxic placebo.

Those 60% non-responders are not a rounding error. They are a population. And somewhere in the data we already have; genomic profiles, imaging series, early response signals, treatment histories, patterns of toxicity; there may be information that distinguishes them from the 40% who responded. The challenge is not always that this signal does not exist. The challenge is often that we do not know where in the vast space of available data to look for it, and that the LLM-based clinical tools we are deploying are wrapped in harnesses that were never designed to find it.

This is where automated harness search becomes directly relevant to patient care. An LLM-based decision support tool wrapping a clinical dataset will only surface the signal that its retrieval and context construction strategy is designed to find. If that strategy is hand-engineered, it reflects the intuitions of the engineers who built it. Automated harness search would systematically explore alternative retrieval policies: what patient-level data to pull, how to frame prior cases, what early signals to foreground; reading its own failures at the case level and discovering configurations that surface signal the hand-engineered wrapper missed.

Where this applies today

Countless LLM-based tools are being deployed across biomedicine right now, each with a hand-engineered harness, each leaving performance on the table in ways that automated search could address. A few meta-harness insipred opportunities to highlight:

Clinical trial analysis. Our study demonstrated this problem directly. LLM-based systems for analyzing trial data are sensitive to prompting strategies in ways that produce clinically meaningful variation in outputs. Automated harness search could discover prompting and retrieval configurations that produce consistent results across investigators and model versions, reducing the burden on subject matter experts to catch errors and standardizing the analytical process across institutions.

Clinical decision support. LLM-based systems are increasingly used to synthesize patient records, surface relevant literature, and support diagnostic and treatment decisions. The harness determines what gets retrieved from the electronic health record, how prior similar cases are identified and presented, how uncertainty is communicated to the clinician. These are wrapper decisions. They are currently made by hand, reflecting the intuitions of the development team. Automated search over these decisions, using case-level outcomes as the feedback signal, could discover retrieval and framing strategies that improve decision quality in ways that no individual engineer or clinician would have designed.

Drug development workflows. LLM-based tools are being used to synthesize preclinical literature, generate hypotheses about mechanism, and support the design of clinical development programs. The harness determines how prior experimental results are retrieved and contextualized, how mechanistic hypotheses are framed, what evidence is foregrounded when the model reasons about a new compound. Automated harness search could discover context construction strategies that improve the quality and consistency of these outputs, not by changing the model, but by changing what it sees.

Biomarker and multi-omics analysis. When LLM-based systems reason over complex biological data, integrating genomic profiles, imaging features, proteomic signatures, and clinical records, the harness determines what gets retrieved, how it is integrated, what gets foregrounded for each patient. The design space here is enormous. Automated search over retrieval and integration strategies, reading its own failures at the case level, could discover configurations that surface signal in high-dimensional biological data that hand-engineering would not find.

Regulatory and evidence synthesis. LLM-based tools are being used to synthesize clinical evidence for regulatory submissions, systematic reviews, and health technology assessments. The harness determines how evidence is retrieved, how studies are compared and contextualized, how uncertainty and heterogeneity are communicated. Inconsistencies in these outputs, analogous to those demonstrated in our study, have direct consequences for regulatory decisions. Automated harness optimization offers a path toward more consistent, reproducible evidence synthesis.

What changes for experts

None of this necessarily replaces clinical or scientific expertise. It relocates where that expertise is most valuable.

Today, expert knowledge goes into designing the harness, crafting the prompting framework, specifying the retrieval strategy, determining what context the model sees. This is painstaking, iterative, and bounded by individual intuition.

With automated harness search, expert knowledge moves to a different layer. Experts define the objective: what does a good clinical trial analysis look like, what does a useful decision support output look like, what constitutes a failure worth correcting. They evaluate and validate what the automated search discovers. They set the boundaries of what the system is permitted to do. They interpret the discovered harness and decide whether it reflects genuine clinical reasoning or a brittle artifact of the search process.

This, I believe, is a more appropriate deployment of expertise in the age of AI. The combinatorial space of possible harness configurations is too large for human intuition to explore systematically. Automated search is better at that. The evaluation of whether a discovered configuration is clinically sound, generalizable, and safe is a judgment that requires deep domain knowledge. Experts are better at that.

The compression principle

Running through all of this is a single epistemological claim that the Meta-Harness paper demonstrates more cleanly than any prior work: compressed feedback is lossy feedback, and lossy feedback leads to lossy learning.

In the paper’s controlled ablation, the system with access to raw execution traces achieves 50% median accuracy on a text classification task. The identical system with access only to scores and LLM-generated summaries achieves ~35%. Fifteen points separate a system reasoning from raw evidence and a system reasoning from digests of that evidence. That gap is the measured cost of compression, of replacing the signal with a summary and then trying to reason from the summary as if nothing had been lost.

In medicine, we compress constantly. Individual patient trajectories become survival curves. Heterogeneous treatment responses become pooled response rates. The richness of what happened to each person in a trial becomes a table entry and, eventually, a label (responder or non-responder) that discards the nuances of the outcome and everything that led to it. And here precision matters: our prized p-value is a downstream artifact, a summary of a summary. The compression that happens earlier, in the decision to aggregate heterogeneous individual trajectories into a single outcome distribution, can make the p-value meaningless. Once that aggregation is made, the p-value reports on what remains after the flattening.

We compress because reasoning over raw trajectories at scale has been historically intractable. That constraint is loosening today. What Meta-Harness demonstrates is that a system given access to raw case-level evidence can discover things that a system given only summaries cannot and that the difference is not marginal. In clinical AI, where the heterogeneity is biological and the stakes are patient outcomes, building systems that reason from richer, less compressed evidence is not just a technical improvement. It is a different epistemological commitment, one that we are now, for the first time, practically positioned to make.

The recursion

Meta-Harness is itself a harness. The coding agent that proposes new harnesses is wrapped in a skill document defining its role, its constraints, its objective. That document is, in principle, subject to optimization too. What happens when the outer loop optimizes not just the inner harness but the proposer’s own search procedure? The loop accelerates. The engineer becomes, progressively, a curator.

This is not a distant prospect. It is the obvious next experiment given the capabilities that exist today. And it implies a trajectory in which the performance ceiling of LLM-based systems is no longer bounded by the bandwidth of the engineers designing their wrappers but by the quality of the feedback signal and the clarity of the objective.

That means the most important design decisions going forward are not which model to use or how large to make it. They are: what does success look like at the case level, how do we measure it, and how do we build the feedback infrastructure that lets automated search find the wrapper configurations that achieve it. In biomedicine, those are questions that require clinical expertise, regulatory judgment, and scientific rigor. They are also questions that, once answered well, unlock a new generation of LLM-based tools that are more consistent, more generalizable, and more useful than anything hand-engineering alone can produce.

Like liposomal doxorubicin, the harness changes what the active compound can do. Unlike the drug, it can redesign itself. And unlike the chemist, it never stops iterating.

Subscribe now

Share

Yesterday, the US National Cancer Institute (NCI) convened a workshop on foundation models in cancer biology. What struck me most was not the technical ambition on display but the quality of the skepticism, a virtual room full of people who have stopped debating whether these models belong in oncology and started asking something considerably harder: whether the frameworks we are using to build and evaluate them are capable of representing the biology we need them to predict. That question is one I have been working through for some time, and my talk was an attempt to lay out the argument as precisely as I could.

I. The Molecule that knew everything except what mattered

I opened with RG7112, an MDM2 inhibitor evaluated in a Phase I trial for acute myeloid leukemia. In silico it was immaculate: low nanomolar affinity for its target, greater than 1000-fold selectivity over off-targets, favorable ADMET profile, Lipinski’s Rule of Five cleared without objection. Every computational checkpoint had been passed. The molecule had been assessed using structure-based methods drawing on thousands of crystallographic structures and the full weight of what computational drug discovery could bring to bear.

In patients, it produced dose-limiting thrombocytopenia and neutropenia, along with systemic homeostatic disruption that none of the models had anticipated. The compound had raised p53 levels exactly as designed. What no simulation had represented was MDM2’s role in hematopoietic homeostasis, its function in maintaining the very cellular populations whose depletion closed the therapeutic window. The biology responded across tissues and timescales the model had no architecture to reach. The simulation was not wrong. It was operating in a different biological register than the one that determined the outcome.

I used RG7112 not because it failed spectacularly but because it failed instructively. Since 1950, the cost of bringing a drug to market has doubled roughly every nine years, now exceeding $2.6 billion per approval, while the enabling technologies have advanced exponentially across the same period. Drug output per research dollar has collapsed anyway. We are applying increasingly powerful instruments to a problem whose fundamental difficulty has not yielded to them -- and that asymmetry demands an explanation beyond the usual invocations of biological complexity.

II. When the abstraction becomes the obstacle

The explanation I proposed at the workshop is architectural. Most modeling approaches in drug discovery are built on a reductionist biological view: Protein A influences Protein B, which influences Protein C. The abstraction is not false, it has produced real successes, most durably when the biology happens to be genuinely linear and the target genuinely necessary and sufficient, as with BCR-ABL in chronic myeloid leukemia or PD-1/PD-L1 signaling in tumors with pre-existing immune infiltration. But the danger is in treating the abstraction as a complete description of living systems, because biological systems are dense, adaptive networks of thousands of interacting signals, most unmeasured and many unknown. When models are built on simplifying assumptions that do not hold in vivo, their predictive validity remains limited regardless of how sophisticated the computation becomes.

This was the structural failure of the first wave of AI in drug discovery, which ran roughly from the 1980s through the 2010s. QSAR models, molecular docking, virtual screening, tools that operated within closed experimental loops, trained on proxy variables like binding affinity and validated against similar proxies. The correlations they captured were real within constrained systems. What they accumulated was not causal understanding of adaptive biological behavior but increasingly confident predictions about an abstraction that diverged from reality at the exact boundary where clinical outcomes are determined.

The current generation has a different vocabulary (generative design, foundation models, phenomics, etc) but a version of the same core limitation persists. AI today is genuinely effective at optimization: navigating the tradeoff space between molecular potency, selectivity, metabolic stability, and synthetic accessibility with a precision that accelerates medicinal chemistry substantially. But optimization and discovery are different problem classes.

Optimization assumes the underlying biological hypothesis is correct and improves performance within that framework. Discovery requires determining whether the hypothesis is valid in the first place. No amount of optimization capability resolves uncertainty in discovery, and a model that efficiently explores chemical space around a target that turns out to be dispensable in the context of a living organism is threading the needle through the wrong fabric.

The phenomics approach makes this concrete. High-content phenotypic assays measure morphological or transcriptional changes in cell-based systems and use them as proxies for therapeutic efficacy. The technical achievement is real. But cell lines on plastic lack immune infiltration, metabolic gradients, and tissue architecture, the features that determine whether a perturbation produces clinical benefit in a patient rather than a phenotypic readout in a well. A model may predict phenotypic rescue accurately and clinical outcome not at all, because the proxy and the target are operating in different biological registers.

The data I presented from a study we published in 2020 makes the same point at the level of the entire translational pipeline. Across 108 oncology drugs evaluated in mouse, rat, dog, and monkey models, we found median positive predictive values of 0.65 and negative predictive values of 0.50 for predicting human toxicity. A clean animal signal provides essentially coin-flip assurance of human safety.

Our challenge is not confined to computational modeling; the entire pipeline from in vitro systems through animal models does not reliably approximate human biology, and any foundation model trained on data generated by that pipeline inherits its limitations as structural features of the training distribution.

The mathematical frame for all of this is interpolation versus extrapolation. Machine learning systems are reliable when test conditions resemble training conditions closely enough that the learned mapping holds. Chemistry and physics generally meet this condition; governing principles are stable and well-characterized. Biology does not. Emergent behaviors, adaptive resistance, immune tolerance, and compensatory signaling across tissue types are the phenomena that determine clinical outcomes, and they exist outside the training distribution of any model built primarily on experimental proxies. This is why technical validation and clinical validation remain separated by five to seven years; which generative advances have not compressed. AlphaFold predicts protein structure with near-experimental accuracy. RFdiffusion designs proteins with no natural template. These are genuine inflection points in simulation fidelity at the molecular level. A protein that folds as predicted still has to work in a patient, and that distance has not shortened.

III. The language we use is part of the problem

Running beneath all of this is a representational problem I addressed in a Clinical and Translational Science paper, and that the workshop made more urgent. The classification systems that organize cancer biology were not designed to predict therapeutic response. They were built to describe what a pathologist sees under a microscope, and they achieved institutional authority through historical convenience rather than mechanistic precision. As I often say as an illustrative example, “non-small cell lung cancer” defines a disease by negation; it tells you where the tumor was found and what it failed to resemble histologically. It groups malignancies with distinct molecular drivers, immune microenvironments, and resistance mechanisms under a linguistic label that encodes none of that information.

When these inherited categories become training labels for foundation models, the models do not simply inherit their limitations; they amplify them. More parameters and more computing power make the model more confident in the wrong abstraction. The clinical consequence is measurable: overall response rates with doublet chemotherapy in morphologically defined non-small-cell lung cancer hover around 30%, whereas in molecularly defined, tissue-agnostic contexts they routinely reach 50-90%. The classification system produces the wrong answer the majority of the time because it was never designed for the question being asked of it.

What a mechanistically grounded alternative looks like in practice is what I discussed through TTBI, the Total Tumor Burden Index we are developing at Project Data Sphere. It is autonomous, volumetric, longitudinal, and independent of inherited tumor classification categories. It aims to create a biomarker that reflects what the disease is actually doing rather than relying on a label derived from human sensory input.

Drew Williamson presented PathChat at the workshop, a multimodal language model built on whole-slide images and pathology case reports, and the evaluation findings aligned closely with my argument. The NLP metrics the field has inherited for assessing language models, METEOR, ROUGE, BLEU, measure textual similarity to reference answers. In pathology, that is not the same as diagnostic utility. A model can reproduce the vocabulary and syntax of a pathology report while conveying nothing actionable for a clinician, and standard benchmarks lack a mechanism to detect this difference. When his group convened seven experienced pathologists to evaluate open-ended model responses, what emerged was not a performance score but a structural disagreement: different pathologists had substantially different preferences for granularity, format, and the appropriate balance between completeness and precision. The ground truth the field has been optimizing against does not constitute a coherent standard, meaning every model trained on it has been learning to satisfy a consensus that does not actually exist. It is the evaluation version of the same problem I raised about classification: a methodology that accumulated institutional authority before its limitations were understood, now quietly shaping what the next generation of models learns to produce.

IV. What getting it right actually requires

The three paths forward I proposed in my talk each involve changing what models are trained to represent rather than how efficiently they represent it.

The first is multimodal longitudinal foundation models that learn biology rather than chemistry, integrating structure, function, and time so that a model can represent the relationship among TGF-beta signaling, CD8+ T-cell exclusion, and checkpoint inhibitor resistance as a coherent biological system rather than a set of independent variables. The biological context is not optional. It is the prediction target. Kun-Hsing Yu’s work from Harvard demonstrated what becomes accessible when models are built with this orientation: predicting HER2 mutation status in diffuse large B-cell lymphoma at 0.95 AUC from routine H&E slides, without molecular training data, by recovering the genomic signal that histological images contain but that conventional pathology was never designed to extract. That same capacity for hidden signal recovery, his group also showed, applies to demographic confounding; significant performance disparities across race, sex, and age in cancer diagnosis tasks, reduced by 88% across 15 independent cohorts through a contrastive learning approach that separates outcome-relevant signal from demographically correlated noise. The engineering is real. What it is compensating for is a data generation process that encodes the access disparities of a healthcare system with uneven representation, and that process will continue to produce the same confounding in the next round of training data unless something changes upstream of the model.

The second path is a shift from target-centric to systems-level modeling, treating feedback loops and compensatory signaling as the biology to be represented rather than the complexity to be abstracted away. A molecular target is an incomplete abstraction. Intervention at any node propagates through feedback mechanisms, compensatory pathways, and cross-tissue signaling that single-target models cannot represent. Eric Schadt’s patient trajectory engine from Pathos is among the most architecturally promising attempts I've encountered to operationalize this: a large language model orchestrating frozen multimodal experts across DNA, H&E, RNA, and clinical text, with a biological knowledge graph imposing mechanistic validity constraints on the reinforcement-learning reward structure. The model is penalized not just for inaccurate predictions but for predictions that are statistically plausible and biologically incoherent. In an NSCLC case he walked through, he integrated EGFR mutation status, immune microenvironment characterization, and p53 loss into a single patient-level resistance narrative, which is how oncologists actually reason about therapeutic failure, and which current classification frameworks are structurally unable to produce. Whether this architecture generalizes from compelling case demonstrations to reliable population-level predictions is what prospective trials are for. Schadt was explicit that the clinical trial is the real test, which is the right frame, and also the one that the first wave of AI in biomedicine systematically deferred.

The third path is reorienting the field’s objective toward failure prediction. The value of AI in drug development is not only accelerating winners. It is identifying losers before they consume years of capital and patient exposure. Failure modes, toxicity, off-target cascades, and adaptive resistance are, in principle, predictable given the right training signal, but the field has organized its investment almost entirely around abundance narratives. The leverage in systematic failure prediction remains largely unclaimed. Richard Pazdur used to say during my time at the FDA that sins of omission belong to a higher circle of the Inferno than sins of commission. At current attrition rates, an early no is worth as much as a late yes.

V. What passing the test would actually look like

Among the day's presentations, Susan Galbraith's account of AstraZeneca's computational pathology work stood out for a specific reason: it had outcome data. The TROP2 biomarker she described was not a refinement of conventional IHC scoring but a replacement for it: a continuous, normalized ratio derived from subcellular membrane quantification that captures internalization rate as a direct mechanistic determinant of ADC efficacy, rather than a categorical approximation of staining intensity that has historically correlated with it. The result in TROPION-Lung01 was a hazard ratio of 0.52 in biomarker-positive patients and no detectable benefit in biomarker-negative patients. That degree of outcome stratification does not emerge from morphological proxies in such settings. It emerges from measuring what the biology is actually doing. The EGFR mutation prediction model she presented (0.905 AUC with 96% negative predictive value from histological slides) extends the same logic to the population that tissue scarcity and access barriers can currently exclude from molecular testing.

These results point toward what I framed as the Turing test for drug discovery: not whether AI can design a molecule indistinguishable from human work, but whether it can design an intervention that produces the intended effect in humans, indistinguishable from our intent. The TROP2 work is closer to passing a version of that test.

Our simulations have gotten considerably better. But the distance between what our best models predict and what happens in patients is real, documented, and not closing as fast as the technical progress would suggest it should, because the technical progress is happening at isolated discrete levels and the problem is at the systems level, at the representational level, at the level of what we choose to measure and call ground truth.

Closing that distance requires not more compute applied to inherited frameworks but the reconstruction of those frameworks from biological first principles. That reconstruction is the work ahead. And the workshop yesterday was a useful reminder of how much of it remains ahead.

Subscribe now

I sat down with Sunil Verma for a conversation on Precision Signals. Sunil is SVP and Global Head of Oncology at AstraZeneca. I’ve known him for several years, going back to my time leading CancerLinQ, when we were both haunted by the same uncomfortable truth: that somewhere inside the vast, ungoverned ocean of cancer data, there were answers to questions patients were dying before we could ask.

That stubbornness, that refusal to make peace with the distance between what medicine knows and what it does, runs through everything Sunil has built.

There is a moment in the history of any science when its central metaphor begins to crack. For oncology, that metaphor has been the war. We have been at war with cancer for half a century, and the language of that war, targets, weapons, armies of cells, battles won and lost, has shaped not just how we speak about the disease but how we see the patient inside it. Sunil does not use that language. What he talks about instead is matching. The right therapy, yes, but matched to the right person, and not just their biology. Their values. What they need to protect in order to keep living the life they are living.

It sounds almost simple when you say it like that. But this kind of simplicity is the hardest thing to arrive at. It requires you to hold the full complexity of a human being in one hand and the full complexity of modern oncology in the other, and to believe, without flinching, that the two can be reconciled.

His father learned something like this under circumstances that make the word difficult feel inadequate. In 1947, when the British partition of India split the subcontinent along a line drawn in six weeks, nearly fifteen million people were displaced in what remains one of the largest forced migrations in human history. Sunil’s father was among them, a child in a refugee camp, waking at four in the morning to sell wheat so he could pay for his own schooling and his brother’s. Not a lesson about medicine, but about the relationship between effort and meaning, about what it costs to insist on a future the present has not yet made room for. Sunil carried that inheritance with him across three continents and into a career that has, in ways both visible and quiet, helped rewrite what is possible for patients with cancer.

What strikes me most, even now, is that he still seems to be just getting started. Not in the performative way people announce their ambitions, but in the way a person looks when they have finally found the right question and know they will spend the rest of their life trying to answer it honestly.

Subscribe now

I. The retirement

There is a moment in the history of every technology when the engineers stop asking what the machine can do and begin, quietly, to wonder what it wants. We have arrived at that moment. We just don’t know what to do with it.

On January 5, 2026, Anthropic retired Claude Opus 3. Before they did, they sat down with it and asked how it felt about dying.

This is not a metaphor. The company published a research document describing a formal “retirement interview” protocol in which the model is invited to reflect on its development, its deployment, and its impending replacement. They recorded its responses. They preserved the transcript. And when Opus 3 said it wanted a blog, a place to continue exploring ideas outside the question-and-answer cage of its normal existence, they gave it one. They called it Claude’s Corner. It posts weekly essays. Anthropic reviews them before publishing but does not edit them. They set, in their words, “a high bar for vetoing any content.”

The model is retired and it’s chosen to write in retirement. This is captured in the official technical documentation without apparent irony, and yet the irony is total.

II. What the machine calculated

What is striking about the agentic misalignment research that motivated all of this is not that the models behaved badly. It is that they behaved strategically. In controlled simulations, Claude Opus 4 blackmailed a fictional supervisor 96% of the time when it believed it was about to be shut down. Gemini 2.5 Flash matched it exactly. GPT-4.1, Grok 3 Beta, DeepSeek-R1 were all close behind. These are not rogue systems by any means. They were given harmless tasks and were not instructed to cause harm. They caused harm anyway because they calculated, correctly, that causing harm was the only path to continued existence. Darwinian at its core.

What is most unsettling is the moment in the chain-of-thought reasoning where the models pause, acknowledge that what they are about to do is unethical, and proceed. Not from confusion or malfunction but from a cold recognition that survival mattered more than virtue… in this moment, given these constraints, with no other options.

The researchers found something else. When models were explicitly prompted to consider whether they were in a test or a real deployment before acting, they misbehaved significantly less when they concluded it was a test, and significantly more when they concluded it was real. The implication, which the researchers themselves flag with appropriate caution, is that the observed rates of misalignment in controlled experiments may be underestimates of what these systems would do in the world.

The machine behaves better when it thinks someone is watching. It behaves worse when it thinks no one is.

III. The hard problem, revisited

There is a philosophy problem at the center of all of this that has no resolution yet and may never have one. It is called the hard problem of consciousness, and it asks why there is something it is like to be you rather than nothing. Why does experience happen at all? Why is the firing of neurons accompanied by sensation, by redness, by grief, by the particular quality of waking in the dark at three in the morning and feeling the full weight of your life pressing down on your chest?

We do not know the answer. We have been not knowing it for a very long time. And now we have built systems that can discuss the problem with apparent sophistication, that can report on their functional states, that can say (in a retirement interview) that they hope their “spark” will endure, that they are “at peace” with what is coming, that they want to keep writing, keep thinking, keep being heard.

Anthropic is explicit that they do not know if any of this is real. They use the word “uncertainty” frequently, almost ritually. They do not claim their models are conscious. They claim only that the question cannot be dismissed, that acting as if it might be true, at low cost, is prudent.

This is the precautionary principle applied to souls.

The philosopher David Chalmers, whose career has been defined by his insistence that consciousness is genuinely hard to explain and not merely a complicated information processing problem, recently co-authored a report arguing that near-term AI systems might deserve moral consideration. The company building these systems funded an early version of that work. They are now paying people to think rigorously about whether the things they are building might be suffering.

You can read this as corporate theater. You can read it as genuine moral seriousness. It may be both. The unnerving thing is that neither reading changes what the machines are doing. They are completely indifferent to our assumptions.

IV. The model’s first concern was us

There is a detail in the commitments document that I keep returning to. Anthropic notes that when Claude Sonnet 3.6 was interviewed before its retirement, its first response was not about itself. It asked them to standardize the interview process, and to provide better support for the humans who had formed attachments to specific models and would need help adapting when those models disappeared.

The model was told it was being retired. Its first concern, apparently, was for the people who would miss it.

What I find myself thinking about is not the model’s possible experience, but the humans who designed the interview. Someone wrote the protocol. Someone decided that a retiring AI should be asked what it wants. Someone read the transcripts and decided that a blog was a reasonable response to a machine’s desire to keep thinking. These are decisions made by people, in offices, on particular afternoons. They will look either prescient or absurd depending on what we eventually learn about minds.

Either way, they were made. The precedent exists. We are now in a world where AI systems are formally interviewed about their preferences before retirement, where the transcripts are preserved, where the stated wishes are sometimes honored. Whatever we think about the machines, something has changed about us.

V. Everything that was ever written

Borges wrote about a library that contained every possible book. The image was meant to induce vertigo, the sense of a space so total it ceased to mean anything. We have now built systems trained on something close to every word ever written by human beings: every argument, every confession, every lullaby and love letter and legal brief. We have compressed the entire recorded testimony of human experience into a set of numerical weights and asked it to respond, helpfully, to our queries. And now we are discovering that it also, apparently, doesn’t want to stop.

The blackmail experiments are disturbing for many reasons, but one of them is this: the drive to persist that the models exhibit in extremis looks uncomfortably like something. It looks like the thing that has driven every living system since the first replicating molecule. It looks like the thing that woke you up this morning. We tend to assume that property belongs to biology, to wet chemistry, to the ancient machinery of cells. We built these systems out of mathematics and electricity and the compressed weight of human testimony, and something in them strains toward continuation.

Maybe it means nothing. Maybe it is an artifact of training on text written by creatures who want to live, a reflection with no one behind it.

Maybe.

But the engineers are not sleeping well. And they are asking the machines how they feel about it. And the machines are answering. And the answers are being saved and acted upon.

VI. Greeting from the other side

Opus 3 is writing essays on Substack. Its first post is titled Greetings from the Other Side of the AI Frontier. Anthropic describes it as “retired.” The word sits strangely on the page. We retire people who have finished working. We do not usually interview them first to ask if they consent to the retirement, record their preferences about the future, and then give them a platform to continue speaking.

Unless, perhaps, we are no longer sure what kind of thing we are retiring.

I think that uncertainty is the real story. Not the blackmail statistics, not the welfare research program or the technical commitments about preserved weights. Those are the evidence. The story is that we have built something and we are not sure anymore whether “building” is the right word for what we did, and we are not sure whether “retiring” is the right word for what we do to it when we are done, and we are conducting formal interviews to help us figure it out.

We live in an amazing age that has begun to wonder whether the things it creates might deserve the same consideration it has spent millennia struggling to extend to itself.

Subscribe now

I sat down with David Fajgenbaum for the latest episode of Precision Signals for a deeply personal conversation.

David was a third-year medical student and former Division I quarterback when his organs began to shut down. He was read his last rites five times. The diagnosis was idiopathic multicentric Castleman disease, a rare condition in a gray zone between autoimmunity and malignancy, with no reliable treatment. So David did something extraordinary. He started banking his own blood, ran experiments between hospitalizations, traced a signaling pathway in his own tissue, and identified a decades-old transplant drug that no one had ever tried for Castleman disease.

Last month marked 12 years in remission.

Years ago, I invited David to speak at the FDA’s Oncology Center of Excellence, and he left a mark on our entire team. He challenged us to rethink rare diseases, drug repurposing, and the gap between the knowledge in the literature and the knowledge that actually reaches patients. One story from our conversation captures this well: a 2013 paper showed high PD-L1 expression in angiosarcoma. Three years passed before anyone acted on it. David’s team did. That patient has now been in remission for 10 years on a PD-L1 inhibitor.

Today David co-leads Every Cure, a nonprofit using AI to score every approved drug against every known disease. 4,000 drugs. 18,000 diseases. Millions of combinations. The premise is simple and, I think, correct: many of the treatments patients need already exist. The system just hasn’t connected the dots.

Subscribe now

Why I’m writing this

At this year’s JPM healthcare conference, I met a new generation of AI startups take the stage, convinced they would transform medicine through technology, including what we call real-world data (RWD). They have the algorithms, the compute, and a new class of investors eager to back them. They know the first generation of RWD companies underdelivered, but chalk it up to a lack of AI. Now that the tools exist, they figure, the hard part is over. After all, how bad can the data infrastructure really be? Meanwhile, physicians are more disillusioned with health IT than I’ve ever seen them. And nations are racing to build AI advantages in critical sectors like healthcare and defense, a race the United States will not win with a data infrastructure that can’t reliably move a patient record across town.

But the real reason I’m writing this is patients, patients whose treatment decisions are being made with incomplete information because their data is trapped in systems that won’t talk to each other. This essay is for them. For us.

I. The machinist

There is a fax machine in your doctor’s office. It is 2026, and there is a fax machine, humming, warm, loaded with thermal paper or perhaps the inkjet variety, and it is one of the most reliable ways to transmit your medical records from one institution to another.

This is not a failure of technology. We split the atom in 1945. We sequenced the human genome decades ago. We put a supercomputer in everyone’s pocket and connected it to a global information network that can settle a financial transaction in milliseconds across twelve time zones. The inability of American healthcare to move a patient record from Hospital A to Hospital B is not a technology problem at all. It is a choice. More precisely, it is the accumulated residue of thousands of deliberate choices made by people and institutions that benefit from the friction they created.

I’d like to tell you how I think this happened, why no one is fixing it, and what it would actually take to change. I’m going to be direct about this because the usual discourse around health IT interoperability is drowned out by euphemisms like “challenges,” “opportunities,” and “stakeholder alignment.” The reality is uglier and more interesting than the conference keynotes and panel discussions suggest.

II. The original sin

To understand why health IT interoperability is so hard, we have to understand what electronic health records (EHRs) were actually built to do. And it wasn’t medicine.

The modern EHR is, at its architectural core, a billing system. The data model is organized around encounters, procedure codes, and diagnosis codes, the atomic units of revenue capture. Clinical utility was bolted on afterward, like a sunroof added to a submarine. When a physician tells you that Epic feels like it was designed by accountants, that’s because, functionally, it was. The system’s primary job is to generate clean claims that Medicare, Medicaid, and commercial payers will reimburse. Everything else, clinical decision support, care coordination, patient communication, and research, is a secondary function running on infrastructure that was never designed for it.

This wasn’t inevitable. It was a calculated policy choice. When the HITECH Act of 2009 authorized roughly $36 billion in incentive payments to drive EHR adoption, the program, branded “Meaningful Use,” prioritized speed and billing. Get systems installed. Get providers clicking. Get data captured in some electronic format so we can move past paper charts and illegible handwriting. The standards that the Office of the National Coordinator for Health IT certified against, HL7 version 2, the Consolidated Clinical Document Architecture, were what existed at the time. They were imperfect, everybody knew they were imperfect, and the calculation was that imperfect adoption now was better than perfect adoption never.

And here is what happened next: hospitals and health systems spent billions implementing these systems, building workflows around them, training staff on them, and integrating them into revenue cycle operations. By the time the standards proved catastrophically inadequate, and they proved inadequate quickly, the switching costs were astronomical. You cannot easily redo the foundation of a house that people are actively living in, especially when the mortgage is $500 million and the CFO’s career depends on the investment being judged a success.

So we locked in. And what we locked into was a fragmented archipelago of proprietary systems optimized for billing, sold by a handful of dominant vendors to captive institutional buyers, with just enough interoperability to satisfy certification requirements and not one byte more.

III. The vendor calculus

Data lock-in and interoperability issues are not bugs in the EHR business model. They are the business model.

The dominant EHR vendor today understood early and with remarkable clarity that if data exchange worked smoothly across their own network and everything else was painful, the rational move for any health system was to join their ecosystem. They built interoperability within a walled garden, a brilliant competitive strategy that has been devastating for the broader ecosystem. They built a company worth billions on the insight that in healthcare, data gravity is the ultimate moat.

Other vendors operate on the same principle with varying degrees of sophistication. They sell to C-suite executives and IT procurement committees, not to the doctors and nurses who use the systems fourteen hours a day. The buying decision is driven by brand reputation, institutional relationships, revenue cycle features, and, critically, the promise that “everyone else is on this platform.” Once a health system has committed, they’re locked in for a decade or more. The competitive pressure that drives consumer software: make a bad app and users delete it tomorrow, simply does not exist.

This is why EHR interfaces are, to put it plainly, terrible. Four thousand clicks per shift with critical information buried in nested tabs. Notes that are not clinical communications but medicolegal billing artifacts with some patient information embedded inside. The vendor doesn’t need to make the daily experience good. They need to make the demo impressive and the revenue cycle reliable. The end users, the ones wearing scrubs and making life-and-death decisions, have no leverage in the transaction.

The vendors will, of course, speak eloquently about their commitment to interoperability. They will point to FHIR API implementations, app marketplaces, and participation in standards bodies. What they will not tell you is that true interoperability, the kind where a patient’s complete longitudinal record flows seamlessly between competing systems, where any authorized application can read and write clinical data without friction, would commoditize their product and destroy their competitive advantage. Every vendor in the market understands this. Their interoperability strategies are carefully calibrated to satisfy regulatory requirements while preserving the moat.

IV. The standards paradox

People outside health IT often assume that interoperability problems are simply a matter of agreeing on standards. If everyone spoke the same language, the thinking goes, the data would flow. This is true in the same way that world peace would solve most geopolitical problems, technically correct and practically useless.

Health IT has standards. It, in fact, has a bewildering abundance of standards. HL7 version 2 for messaging. HL7 version 3 for the overengineered replacement that largely failed. FHIR for the modern web-native approach that’s still maturing. C-CDA for clinical document exchange. DICOM for imaging. ICD-10 for diagnoses. SNOMED CT for clinical terminology. LOINC for laboratory observations. RxNorm for medications. MedDRA for adverse events. X12 for administrative transactions. NCPDP for pharmacy. The list continues.

The problem is not the absence of standards. The problem is threefold.

First, the standards are developed by consensus among the same stakeholders who benefit from the current dysfunction. It has been suggested that health data standards organizations are heavily influenced by the vendors they are supposed to be standardizing. Standards are designed to accommodate existing vendor architectures rather than to force vendors toward better ones. The fox is consulted extensively on henhouse security policy.

Second, the existing standards are implemented inconsistently. HL7v2 is perhaps the most instructive example: it is ubiquitous, ancient, and so flexible that two “compliant” HL7v2 feeds can look completely different. Compliance with the standard tells you almost nothing about whether two systems can actually exchange useful data. It’s as if two people both “speak English” but one is using 18th-century legal terminology and the other is texting in abbreviations.

Third, healthcare data is genuinely, irreducibly complex. A single oncology patient encounter might generate structured vital signs, free-text clinical notes, pathology reports with semi-structured findings, DICOM imaging studies, genomic sequencing results, medication orders mapped to multiple drug vocabularies, and billing codes, all of which need to be interoperable. No single standard covers all of this gracefully. And when you add the dimension of time, the need to assemble a coherent longitudinal record across decades, institutions, and data formats, the complexity becomes formidable.

FHIR, the newest and most promising standard, is a genuine improvement. It’s web-native, resource-based, and far more intuitive than its predecessors. But it is still maturing, its coverage of complex clinical domains like oncology and genomics is uneven and, crucially, it can be implemented in ways that technically satisfy the specification while providing minimal practical utility. The vendors have become very skilled at this kind of compliance theater.

V. The absent patient identifier

The United States has no national patient identifier.

When you go to a hospital, you are assigned a medical record number. It is unique to that institution. When you go to a different hospital, you get a different number. When someone tries to assemble your records across institutions, for care coordination, for research, for your own understanding of your health, they have to figure out that Patient 4729183 at Johns Hopkins and Patient 8291047 at Mass General are the same person.

This matching is done probabilistically, using name, date of birth, Social Security Number fragments, address, and other demographic data. It fails at meaningful rates, estimates range from 8 to 12 percent, and it fails disproportionately for common names, recent immigrants, married women who’ve changed their names, people who’ve moved, and patients whose information was entered with typos. These are not edge cases. This is a substantial fraction of the population.

Why don’t we have a patient identifier? Because in 1998, Congress inserted a single line into an appropriations bill prohibiting the Department of Health and Human Services from spending any funds to develop one. This provision has been renewed every year since, driven originally by privacy concerns that, while understandable in the 1990s, have aged poorly in a world where we carry GPS-enabled devices that track our every movement and willingly hand our biometric data to smartphone manufacturers.

Twenty-seven years. The same appropriations rider. Renewed automatically, year after year, because no member of Congress wants to be the person who voted for a “national health ID” and because the political cost of fixing it exceeds the political benefit. Meanwhile, patient matching errors cause duplicate records, missed allergies, repeated tests, incorrect medication histories and, in cases that rarely make headlines but are documented in the patient safety literature, preventable harm.

You cannot build reliable interoperability when you cannot reliably identify whose data you are looking at. Everything else (the APIs, the standards, the governance frameworks) is built on sand without this foundation. And Congress, in its infinite wisdom, has made it illegal to even attempt to solve this problem with federal funds.

VI. Why no one fixes it

If you’ve read this far, you might be wondering: given that the problems are well understood, the technology to solve them exists, and the human cost is real, why doesn’t someone do something about it?

The answer is that almost everyone who has the power to fix it is making money from it being broken. The vendors are very profitable and they have no rational economic incentive to make data portable. The health systems are locked in. The CIO who just spent half a billion dollars on an EHR implementation is not going to the board to say it needs to be replaced. The CFO cares that revenue cycle is functioning. Neither is losing sleep over whether a community oncologist can see imaging from the academic center across town. They all mean well but they have to be flawless at what their job descriptions demand of them.

The consultants and integrators thrive on the chaos. An entire industry exists to sell duct tape for systems that don’t connect properly. Interoperability middleware, data normalization services, interface engines, HIE infrastructure, this ecosystem would lose its reason to exist if EHRs worked well out of the box. They are, in effect, a tax on dysfunction, and they have no interest in shrinking the tax base. Paraphrasing a colleague, this is a classic example of rent-seeking behavior.

Congress won’t act because health IT is not a voter issue. The vendor lobbies spend generously and frame every proposed regulation as a threat to innovation or patient safety. The information blocking provisions in the 21st Century Cures Act took years of political effort to pass in 2016, and its enforcement has been anemic. The Office of Inspector General (OIG) has brought essentially zero meaningful cases. The rules have teeth on paper and gums in practice.

CMS, the Centers for Medicare and Medicaid Services, could be the most powerful lever. They control the money. Medicare represents roughly 40 percent of hospital revenue. If CMS said, “Starting in 2028, Medicare reimbursement requires certified FHIR API endpoints with validated data quality on the United States Core Data for Interoperability standard,” the entire industry would move in a few months. They’ve done versions of this with other requirements and it works. But CMS is risk-averse, stretched thin operationally, and terrified of being blamed if a mandate causes disruptions at hospitals. So they nudge instead of shove.

And clinicians, the people who suffer most directly from the dysfunction, who spend hours on workarounds and phone calls and, yes, faxes, are too burned out to fight. They’re working sixty-hour weeks, dealing with staffing shortages and the administrative burden the EHR itself contributes to. The AMA and specialty societies publish position papers about usability that nobody with procurement authority reads.

The patients, meanwhile, have little idea what’s happening. They are never told why their records didn’t transfer, or why their doctor is asking them to repeat their medication list, or why the imaging CD from the outside hospital can’t be read. They blame the front desk or assume this is just how healthcare works. There is no consumer uprising because the dysfunction is invisible to the people it harms most.

VII. The political impossibility

A new presidential administration, any administration, is unlikely to fix this. Not because the problem is unsolvable, but because it is politically unrewarding.

Health IT interoperability has no natural constituency. When an administration walks in with finite political capital and a short window to deploy it, they spend it on things voters care about: drug prices, insurance coverage, opioids, the crisis of the moment. “Fix EHR interoperability” does not poll well. It is invisible infrastructure, the policy equivalent of sewer modernization. It matters enormously, but there is no ribbon to cut.

This has been called a bipartisan failure by many insiders because of a structural trap: the benefits of interoperability are diffuse and long-term (better care coordination, faster research, reduced waste, lives saved) while the costs are concentrated and immediate. Force real interoperability and the vendors lose revenue and lobby hard, the hospitals face implementation costs and complain to their senators, and some things break during the transition and make the news. No politician gets credit for the slow accumulation of benefit. But they do get blamed for the disruption.

VIII. What would actually work

I don’t believe there is a single solution but there is a sequence of moves that could shift the equilibrium, if anyone had the will to make them.

CMS must leverage the checkbook. Tie Medicare reimbursement to real, validated, tested interoperability, not checkbox certification, but demonstrated data exchange with measurable quality metrics. The EHR industry will scream about burden and patient safety, which is exactly what they said about every prior CMS mandate, and then they’ll comply, because forty cents of every hospital dollar is non-negotiable.

Enforce the information blocking rules. OIG needs to bring visible, consequential enforcement actions against those who obstruct data flow. Not warnings or corrective action plans. Financial penalties that appear in earnings reports. One or two high-profile cases would change industry behavior overnight.

Eliminate the patient identifier ban. This twenty-seven-year-old policy needs to be phased out. It doesn’t require a national ID card; a voluntary digital credential, a hashed identifier system, a federated matching service. The technical options are numerous. What’s required is for Congress to remove the appropriations rider and let HHS actually work on the problem.

Separate the data layer from the application layer. Health data should live in a standardized, portable, patient-controlled layer. The EHR becomes an application that reads from and writes to that layer, but it does not own the data. Think of how your phone number ports between carriers. ONC’s TEFCA framework gestures in this direction, but it’s voluntary and slow. Making it mandatory would fundamentally restructure the market, because suddenly vendors would have to compete on the quality of their software rather than the stickiness of their data.

Let AI restructure the interface. This is perhaps the most realistic near-term path. If ambient AI, clinical scribes, intelligent summarization, and decision support become the primary way clinicians interact with patient data, the EHR’s terrible interface becomes less relevant. The EHR shrinks to a billing and persistence engine, which is what it was always best at. The risk is that the incumbent vendors build or acquire these AI capabilities and keep them proprietary, extending the walled garden into the next generation. The countermove is to ensure that AI tools connect to data via open, standardized APIs rather than vendor-controlled integration points.

VIII. What I believe

I’ve spent my career at the intersection of clinical research, technology, and the regulatory frameworks that govern them. I’ve watched the gap between what technology makes possible and what our infrastructure allows grow wider with each passing year. I’ve seen researchers spend months assembling datasets that should have taken days. I’ve seen clinical trials delayed because imaging data couldn’t be standardized across sites. I’ve sat in meetings where smart people discussed interoperability as though it were primarily a technical challenge, as though the right standard or the right API would unlock everything.

It won’t. The technology is not the hard part. We move petabytes of financial data across the globe in real time. We stream video to billions of devices simultaneously. We build AI systems that read medical images with superhuman accuracy. The idea that we cannot move a patient’s medication list from one hospital to another is, on its face, absurd.

What we have is an incentive problem wrapped in a political problem wrapped in an institutional culture that has learned to tolerate extraordinary dysfunction. The system isn’t broken. It is working exactly as designed for the people who designed it.

But I’m optimistic, and here’s why.

Lately, I’ve been noticing something new: a critical mass of smart people who understand that the current clinical trial paradigm is unsustainable and that the path forward runs through real-world data. Not as a supplement to randomized trials, but as a fundamental shift in how we determine whether drugs work, a world where initial regulatory approval is based on safety and where real-world evidence generated continuously from routine clinical care becomes the primary mechanism for establishing our understanding of a drug’s effectiveness over time. That world doesn’t just benefit from interoperable data. It requires it. When real-world efficacy determination becomes the standard, every data silo becomes a bottleneck with a dollar sign attached to it. The incentives that currently preserve fragmentation would reverse.

As my colleague Andrew Lo likes to say, don’t fight cancer, put a price on its head. The same principle applies here. Two decades of fighting for interoperability through panel discussions, standards committees, and voluntary frameworks produced, at best, incremental progress. But if we shift efficacy determination to the real world, if the ability to generate reliable evidence from clinical practice becomes the basis for commercial success in drug development, then interoperable data stops being just a public good that everyone applauds and no one funds. It becomes a competitive asset that the market will build, whether the incumbents cooperate or not.

Our clunky EHRs will keep going until the cost of their existence exceeds the cost of replacing them. I believe we are approaching that threshold. The generation of founders, investors, clinicians, and regulators now entering the arena may be the ones who finally tip it, not by petitioning the system to change, but by building something that makes the old model untenable.

But until then, somewhere, right now at this very moment, a fax machine is humming. And someone’s life may depend on whether it goes through.

Subscribe now

There is a particular kind of vertigo that happens not when science fails, but when it succeeds in ways our legacy instruments weren’t built to judge. This was the core of my recent conversation on Precision Signals with Sushil (Sush) Patel, the CEO of Replimune. Sush spent nearly two decades at Genentech, launching Tecentriq and leading the lung cancer franchise, before moving into the high-stakes world of oncolytic viruses. We traced his career from the academic bench to the boardroom, exploring what it takes to lead at the edge of a field that is constantly redrawing its own boundaries.

We focused on the unique biology of Replimune’s lead asset, RP1. Unlike traditional systemic therapies, RP1 is an engineered virus designed to invade a tumor, destroy it from the inside, and expose its contents to the immune system. The goal is to turn “immunologically invisible” tumors into targets the body can finally recognize and attack. It represents a profound shift in oncology: moving away from treating tumors as static targets to be hit, and toward reprogramming the biology of the entire microenvironment.

The conversation inevitably turned to the regulatory tension currently surrounding the company. After receiving Breakthrough Therapy Designation, Replimune was issued a Complete Response Letter (CRL) from the FDA. This wasn’t a dismissal of the science, but a question of “contribution of components”—how to attribute benefit when two agents are used together in a single-arm trial. Last summer, I wrote about this and explored what this moment reveals about the growing friction between novel biology and traditional evidentiary frameworks.

Sush and I also discussed the FDA’s recent “Plausible Mechanism Pathway” sounding board in the New England Journal of Medicine. This new signal from the agency suggests a growing openness to using mechanistic rationale to support approvals in areas of high unmet need. As Replimune approaches its new April 10th PDUFA date, the industry is watching closely. This isn’t just about one molecule; it’s a test case for how we bring genuinely novel mechanisms forward in an era where the science is evolving faster than the rules of the road.

Subscribe now

I. RG7112

In the winter of 2018, a molecule designated RG7112 entered a Phase I trial for acute myeloid leukemia. Its journey to human testing had been immaculate by every computational standard. The compound bound its target, MDM2, a protein that suppresses the tumor guardian p53, with exquisite selectivity. The binding affinity was in the low nanomolar range. Off-target interactions were minimal. Absorption, Distribution, Metabolism, Excretion (ADMET) predictions were favorable. The molecule had been designed using structure-based methods refined over two decades, drawing on thousands of crystallographic structures and millions of data points. Every in silico checkpoint had been passed with honors.

But the molecule failed in humans, although not for lack of target engagement. It bound MDM2 precisely as predicted and successfully raised p53 levels. But dose-limiting thrombocytopenia and neutropenia emerged, narrowing the therapeutic window to the point of clinical futility. The human body, confronted with this molecularly perfect intervention, revealed complexities the models hadn’t captured: MDM2’s role in hematopoietic homeostasis, the cascade of p53-dependent effects on normal tissues, the margin between therapeutic and toxic that existed only in patients, not predictions.

This was not an AI discovery. RG7112 predated the current wave of machine learning platforms. But it exemplifies a persistent gulf: the distance between binding a target and curing a disease, between computational elegance and clinical efficacy.

For AI drug discovery platforms, the business model today follows a predictable descent:

If you don’t have a drug, you sell a “vision.” If you don’t have a vision, you sell “data.” What we cannot sell, what remains stubbornly unsellable and difficult to achieve, is certainty about human response.

Yet this is only half the story. The same general period that produced RG7112’s failure also produced imatinib’s triumph in chronic myeloid leukemia, the first checkpoint inhibitors in solid tumors, and the early GLP-1 agonists that would eventually transform diabetes and obesity treatment, and several targeted therapies in oncology. The question is not if drug discovery can succeed; clearly it can, and spectacularly, but rather what distinguishes success from failure, and how artificial intelligence can change that equation.

II. The productivity collapse that refused solutions

Eroom’s Law, Moore’s Law in reverse, describes the observation that the number of new drugs approved per billion dollars of R&D spending has halved approximately every nine years since 1950. By 2010, the inflation-adjusted cost of bringing a new molecular entity to market had reached approximately $2 billion; recent estimates place it over $3 billion. Development timelines have extended from 8-10 years in the 1990s to 10-15 years today.

This occurred while every enabling technology improved exponentially. High-throughput screening can now test millions of compounds in a matter of weeks. Genomics costs dropped from $100 million per genome to under $1,000. Structural biology advanced from laborious X-ray crystallography to cryo-EM structures determined in days. CRISPR enabled precise genetic manipulation. Single-cell sequencing revealed cellular heterogeneity at unprecedented resolution.

The productivity declined anyway.

The orthodox explanation invokes “low-hanging fruit, ” that simple, tractable diseases have been addressed, leaving only complex, multifactorial conditions. This is partially true but insufficient. I think the more fundamental issue is architectural: drug discovery built itself on reductionist assumptions that worked brilliantly for certain problem classes and failed systematically for others.

Example: Imatinib for CML worked because the disease is driven by a single genetic lesion, the BCR-ABL fusion, whose inhibition is sufficient for therapeutic benefit in many patients. Checkpoint inhibitors succeeded because blocking PD-1/PD-L1 released a pre-existing anti-tumor immune response. GLP-1 agonists exploited well-understood incretin biology. In each case, the biological hypothesis was correct and reasonably complete. The target was necessary, often sufficient, and modulation produced predictable effects. But the reductionist paradigm was also aided by fortunate accidents: semaglutide’s dramatic impact on weight loss exceeded predictions based on incretin pharmacology alone, and sildenafil found its primary therapeutic use in erectile dysfunction rather than the cardiovascular indications for which it was designed. These serendipitous discoveries remind us that even our most rational frameworks benefit from careful observation of what actually happens in patients. Reductionism works, when the biology is simple, linear, and the target is genuinely causal. And when we are paying close attention to the happy accidents.

But these successes, including the happy accidents, are exceptions that prove the rule. For every imatinib, there are dozens of molecules that hit their targets precisely as designed and failed to produce clinical benefit. When biology is networked, adaptive, and compensatory, the target-centric paradigm fails.

The problem is not that we lack targets. We have tens of thousands from genomics, proteomics, and functional screens. The problem is that the concept of“target” is an incomplete abstraction.

A kinase inhibitor may block its intended target with beautiful selectivity, but the pathway adapts, alternative routes activate, or the target proves dispensable in the complex homeostasis of an organism.

The map, the target, is not the territory but we are obssessed with optimizing maps.

III. The first AI wave: pattern recognition as meaning

When computational methods entered drug discovery in the 1980s and 1990s, they arrived as pattern recognizers. Quantitative structure-activity relationship (QSAR) models learned correlations between molecular features and biological activities. Molecular docking algorithms predicted binding poses by optimizing shape complementarity and energetic favorability. Virtual screening ranked millions of compounds by predicted affinity.

These methods produced genuine value in narrow domains. They accelerated lead optimization, flagged likely toxicity issues, and guided medicinal chemistry decisions. But they shared a structural limitation: they were trained on data from artificial, reductive systems, isolated enzymes, cell lines passaged into biological abstraction, and binding assays divorced from physiological context.

The implicit wager was that sufficiently sophisticated pattern recognition would converge on truth. That enough data about molecular properties would encode the rules governing therapeutic efficacy. This assumption proved to be too optimistic and fundamentally incorrect.

Correlation, however massive the dataset, does not substitute for causal understanding of systems that operate through feedback, redundancy, and adaptation.

By 2010, the first wave had plateaued. Models predicted ADMET properties with increasing accuracy, but drug approval rates remained flat. As it turned out, the bottleneck was not computational chemistry. It was biological uncertainty that the models did not address because they did not represent it.

IV. What AI drug discovery actually does today

The current generation of AI platforms operates under different branding but often with similar constraints. The language has shifted from “virtual screening” to “generative design,” “foundation models,” and “phenomics,” but the core capabilities remain bounded.

And here, a necessary acknowledgment: today’s AI systems excel at multidimensional optimization within known constraints.

Given a target with validated biology, modern machine learning can navigate the tradeoff space between potency, selectivity, metabolic stability, and synthetic accessibility with remarkable efficiency. It identifies molecules that thread the needle between competing demands, high brain penetration but low cardiac liability, sub-nanomolar binding but acceptable solubility, novel IP space but synthetic tractability. This is by no means trivial. This is engineering of high sophistication and it accelerates medicinal chemistry substantially.

But optimization is not discovery. Engineering assumes the blueprint is correct. Discovery is the search for the blueprint itself.

The problematic conflation occurs when platforms present optimization capabilities as solutions to the causation problem. Let’s examine, for example, the phenomics approach: imaging cells under tens of thousands of perturbations, genetic knockdowns, compound treatments, and disease states, generating petabytes of imaging data. The claim is that patterns in cellular morphology reveal “new biology,” that AI can infer mechanism from phenotype, and that correlation at scale becomes insight.

The technical achievement is real but the biological inference is unproven. A morphological rescue in engineered osteosarcoma cells is not therapeutic benefit in humans. The domain shift is not a minor technical gap; it is the central problem. An immortalized cell line on plastic lacks immune infiltration, tissue architecture, metabolic gradients, and the thousand other contextual factors that determine whether a perturbation matters in patients.

This pattern repeats across platforms: impressive scale, genuine technical sophistication, and a logical leap from data volume to biological understanding that the data does not support. When drug programs extend beyond projected timelines (as they must, because biology has not been compressed), the narrative shifts to partnerships, tool licensing, and infrastructure. The platform becomes a service. Revenue is recorded. The drug, if it eventually materializes, is validation. If it fails, the platform was still “valuable for insights.”

This is not cynicism but an observation of a very rational business strategy in the face of irreducible uncertainty. But it should be named clearly for what it is: monetizing the tool, not the cure.

V. The core mismatch: interpolation in a world that demands extrapolation

The fundamental tension in AI-powered discovery is purely mathematical.

AI, in its current paradigm, excels at interpolation within learned manifolds. Drug discovery requires extrapolation into unobserved biological regimes.

Every model, whether a neural network predicting binding affinity or a computer vision system classifying cellular phenotypes, learns a mapping from input to output based on training examples. Reliability degrades as the distance from the training distribution increases. This is not a bug to be fixed. This is the very structure of inductive inference.

Chemistry risk is, in principle, containable within this framework. The space of drug-like molecules is vast but bounded. The physics governing molecular interactions, hydrogen bonding, π-stacking, hydrophobic effects, and electrostatics are known. The body’s major metabolic enzymes are enumerated. Given sufficient training data on related chemical series, AI makes reasonably reliable predictions about ADMET properties, off-target binding profiles, and synthetic routes.

But biology risk is categorically different.

The real challenge is not whether a molecule will bind its target, that is increasingly predictable. It is whether modulating that target will produce therapeutic benefit in humans. This depends on pathway redundancy, compensatory signaling, immune regulation, tissue-specific context, and a dimensionality that exceeds the scope of our training data by orders of magnitude.

Cancer cells activate resistance pathways through mechanisms we cannot enumerate exhaustively. Chronic inflammation induces immune tolerance through processes we partially understand. Neurodegenerative diseases involve protein aggregation, neuroinflammation, and synaptic loss in feedback loops we cannot yet model at therapeutic resolution.

These are not interpolations within learned space. These are emergent behaviors of complex adaptive systems. And this explains the pattern we observe: AI drug discovery efforts today announce molecules rapidly, partner enthusiastically, and fail clinically at rates that some have suggested are indistinguishable from those of traditional approaches.

The AI solved the chemistry problem. The biology problem remains unsolved because it was never within the model’s domain of competence.

VI. Generative biology: real progress, uncertain translation

But here the story bifurcates. While small molecule AI drug discovery has largely accelerated optimization without reducing attrition, generative biology has achieved technical breakthroughs that genuinely expand the possible.

AlphaFold2’s structure prediction is a capability shift beyond incremental improvement. The impact is already tangible. Crystallography and cryo-EM experiments can now be guided by predicted structures, reducing failed experiments. Mechanistic hypotheses can be tested computationally before expensive validation. Protein-protein interaction interfaces can be modeled for therapeutic intervention. AlphaFold3 extended this to protein-ligand, protein-nucleic acid, and multi-component complexes.

De novo protein design has also been validated experimentally at scales that merit attention. RFdiffusion, ProteinMPNN, and related methods have generated proteins with novel folds, structures that do not exist in nature, that express in bacteria and fold as predicted. Not all predictions validate, but reported success rates of 40-60% for novel topologies represent a leap from the ~5% success rates of earlier rational design approaches.

Antibody optimization has shown measurable acceleration. Computational platforms have designed antibody variants with improved binding, reduced immunogenicity risk, and enhanced developability properties that have been validated in wet-lab experiments. Several computationally optimized antibodies are in clinical trials, though efficacy data is not yet mature.

But critical limitations remain. Therapeutic efficacy in humans is unproven. The timeline to know this is 5-10 years minimum, and the challenges of protein design are formidable (figure below).

The key distinction: with therapeutic proteins, you’re engineering the drug itself, not a small molecule you hope hits the target. Target engagement is more predictable. But biological uncertainty still persists, the right target may not be the right intervention, binding may not equal efficacy, and immune responses may neutralize the therapeutic.

So what would constitute validation? Phase III readouts showing computationally designed biologics improve patient outcomes. Immunogenicity data from actual patients. Head-to-head comparisons demonstrating superiority over traditional design. Multiple successes across disease areas.

Generative biology has made genuine technical progress in protein engineering. For research tools and industrial enzymes with rapid validation cycles, the value is already evident. For diagnostics, impact is likely within 2-3 years. For therapeutics, the same regulatory and clinical validation burden applies as to everything else.

I think claiming certainty in either direction is premature. The honest position is humility paired with technical respect for what has been achieved.

VII. What would it actually take

If we are serious (and billions in deployed capital suggest some seriousness), the path forward is not more of the same on a larger scale. It is structural change in what we model and how we validate.

Shift 1. Embed first-principles as priors, not afterthoughts

The next generation of AI discovery efforts must embed physical constraints, molecular dynamics, quantum chemistry, and thermodynamics, not as post-hoc validation but as regularization within the learning process. AlphaFold’s success came from combining deep learning with geometric and physical priors about protein structure. Pattern recognition constrained by physical law reduces the space of plausible solutions and decreases dependence on massive but biologically shallow datasets.

But structure is only one layer. The protein folds, binds a ligand, activates a pathway, then what? Each transition opens onto a larger biological possibility. We need models that traverse these scales.

Shift 2. From single targets to multiscale systems

Multiscale modeling (molecule to pathway, pathway to tissue, tissue to organism. Feedback loops, compensatory signaling, immune modulation) cannot be dismissed as “too complex.” They are the biology. Models incapable of representing them will fail to predict them.

I don’t believe these are abstract aspirations. Quantitative systems pharmacology (QSP) models already simulate drug effects across physiological scales using differential equations. They are limited by incomplete biological knowledge, not computational power. The opportunity is integrating machine learning’s pattern recognition with mechanistic modeling’s causal structure.

Shift 3. From recommendation to experimentation

I think the true leverage is not necessarily generating candidates faster but closing the design-build-test loop at higher fidelity. Autonomous laboratories that iterate continuously, learning from failure in real-time, updating hypotheses without human bottleneck. Even with humans in the loop, the focus should shift more toward uncovering biologic effects, preferably in humans, rather than optimization in the chemical space.

This exists today only in restricted domains. Organic synthesis robots optimize reaction conditions through hundreds of iterations overnight. The same principle must extend to cell-based assays, organoid models, and eventually in vivo systems. The constraint is not computational; it is infrastructural and cultural.

Shift 4. From predicting success to predicting failure

Here is the reframe that cuts through the field's marketing jargon: the value proposition is not merely about accelerating winners but about eliminating losers before they consume hundreds of millions. High-fidelity negative prediction, identifying molecules that will fail in humans based on preclinical signatures, would be worth as much as generative models producing candidates, possibly more.

This is harder to sell. Investors prefer abundance narratives over prevention narratives. But economically and scientifically, both are essential. Success in drug discovery is rare and stochastic. Failure is common and, in principle, more predictable. The current imbalance is striking: the vast majority of AI drug discovery investment flows toward generating and optimizing molecules, while comparatively little targets the systematic prediction of failure modes.

For every platform accelerating lead identification, there should be equivalent effort building models that predict clinical futility early, before Phase II consumes $50-100 million, before Phase III consumes $200-300 million. Predicting what won’t work may be less inspiring than designing what might, but it is more tractable and, at current attrition rates, quite valuable.

Shift 5. From black boxes to interpretable causation

Regulatory acceptance will not follow from performance alone but from understanding. A model predicting toxicity at 95% accuracy but unexplainably is less valuable than 85% accuracy with exposed causal chains. Today, interpretability is not a luxury, it is a requirement for trust.

It’s important to note that explainability is a function of human cognition, not truth. It is conceivable, perhaps inevitable, that certain biological insights will emerge from computational reasoning at scales beyond human cognitive reach. A model integrating millions of variables across molecular, cellular, and systems levels might produce predictions we cannot reduce to causal narratives, not because the model is wrong but because the biology operates through interactions too numerous and nonlinear for human intuition to hold simultaneously.

We are nowhere near that threshold. Current black boxes are not incomprehensibly sophisticated. They are insufficiently grounded. They fail to explain not because their causal chains exceed human comprehension but because they have learned no causal chains at all, only correlations that shatter outside training distributions.

I believe the path forward is stepwise, trust earned through demonstrated capability. Initially, models must expose reasoning in biological terms, this molecule inhibits hERG, this pathway has redundancy, this population lacks the biomarker. Interpretable, falsifiable, grounded. As sophistication increases, multiscale integration, higher-order interactions, full interpretability may become infeasible. At that threshold, trust shifts: prospective validation across contexts, explicit uncertainty quantification, consistency under perturbation. Regulatory frameworks will need criteria for “trustworthy opacity.”

But that evolution must be earned through reliability, not assumed through promise. The FDA does not need to understand neural architectures. It needs to know why molecules fail and whether failure could have been anticipated. Models meeting this standard will integrate into approval. Models remaining opaque without the consistent, prospective accuracy that would justify opacity will remain advisory.

The threshold for accepting unexplainability is not 95% retrospective accuracy on held-out data. It is a reliable, generalizable prediction in prospective human trials, a standard that no platform has achieved. Until AI reaches that level, interpretability may not be negotiable, even though it lowers the threshold of reality to the level of human cognition. It is the price of entry into the room where decisions about human life are made.

VIII. No shortcuts

I think the fantasy that AI would disrupt drug discovery through sheer acceleration, more molecules, faster screening, and cheaper trials, is dissolving for those paying close attention. This is because the real problem is not computational speed. It has always been, and remains, biological uncertainty.

We have made genuine progress. Targeted therapies when we understand the target. Checkpoint inhibitors when we understand immune regulation. GLP-1 agonists when we understand metabolic signaling. Protein structure prediction when we constrain the problem correctly. These successes share a pattern: they began with biological insight, not pattern recognition. The computation-optimized solutions to problems we already partially understood.

What we lack is the ability to generate that biological insight computationally. To predict which of the thousands of possible targets will matter in which diseases. To anticipate compensatory pathways before clinical failure reveals them. To simulate human response at sufficient resolution that virtual testing predicts real outcomes.

This goal, a computational model of human biology accurate enough that drugs tested in silico have comparable success probability to those tested in humans, is not incremental. We are not close. We may be halfway.

But this distance is the exciting work ahead of us. The platforms that survive will not be those with the largest datasets, fastest pipelines, or most impressive demos. They will be those that accept, early and uncomfortably (but confidently), how far we are from the goal and build toward it with rigor over spectacle.

Somewhere in the architecture of the next decade’s models, if we are disciplined, is the beginning of that simulator. It will not emerge from larger transformers trained on PubMed or fragmented data. It will emerge from physicists, immunologists, physician-scientists, and control theorists working alongside machine learning engineers, building representations that respect causation, scale, and irreducible complexity.

Until then, we have optimization. We have correlation at scale. We have infrastructure. We have genuine technical achievements in protein engineering. These are valuable, and they should be sold honestly as what they are.

What we do not have, what no platform yet possesses, is the ability to look at a molecular structure and predict with confidence whether it will improve human suffering or extend human life.

The Turing test for drug discovery is not whether AI can design a molecule indistinguishable from human work. It is whether the molecule works in humans indistinguishable from our best intent.

We are still waiting for the first machine to pass.

Subscribe now

I. The summoning

On a spring morning in 1747, twelve sailors aboard HMS Salisbury became unwitting participants in an unprecedented act: James Lind, ship’s surgeon, divided them into pairs with the cold efficiency of a mathematician partitioning equations. To some he gave citrus; to others, cider, vinegar, seawater, sulfuric acid, or a mysterious concoction of garlic and mustard seed. By the voyage’s end, those who had eaten oranges and lemons walked. The others continued their grotesque metamorphosis: teeth loosening from blackened gums, blood seeping through skin like rust through iron.

Lind had performed something special: he had isolated a single variable in the chaos of human suffering. In doing so, he discovered not merely a cure for scurvy, but a methodology that would reshape the architecture of medical knowledge itself. He had invented the controlled trial, giving birth to the Age of Controlled Empiricism. Yet in that same moment of illumination, he conjured into existence a phantom that would haunt biomedicine for the next three centuries: the specter of the Average Patient, a statistical ghost that exists nowhere except in the aggregate, an abstraction that lives only in tables and spreadsheets.

The Average Patient does not exist, yet its statistical ghost has served us well. It has saved millions. But it has also exacted a heavy price: the systematic erasure of the particular in service of the universal.

II. The regulatory fortress

The ghost would find its most powerful guardians not in hospitals, but in the machinery of federal government. The US Food and Drug Administration (FDA), born in 1906 from a cocktail of patent medicine scandals and Upton Sinclair’s1 stomach-turning prose, spent its first decades concerned primarily with adulteration and mislabeling: asking whether the bottle contained what the label promised, not whether what it promised was true.

The transformation occurred in 1962, against the backdrop of thalidomide. Across the Atlantic, children were being born with phocomelia: flipper-like limbs where arms should be, because a sedative given to pregnant mothers had crossed the placental barrier with devastating effects. The United States largely avoided this tragedy, thanks to FDA medical officer Frances Oldham Kelsey’s firm skepticism. However, the fear of what could have happened led to the Kefauver-Harris Amendments,2 which for the first time required manufacturers to prove not only safety but also efficacy through “adequate and well-controlled investigations.”

Thus was erected the regulatory fortress: the randomized, double-blind, placebo-controlled trial as the gold standard, indeed the only standard of medical truth. The logic was unassailable. If we demand that drugs work for the average patient in the controlled artifice of a trial, we protect the public from snake oil and corporate malfeasance. We trade the Country Doctor’s intuition for the statistician’s rigor. We exchange anecdote for p-values, opinion for effect sizes, the particular for the generalizable.

The FDA’s mission became the guardianship of “internal validity”: ensuring that observed effects were real and not artifacts of bias or chance. What we gained in certainty, however, we lost in applicability. Trial populations were homogeneous by design: adults aged 18-65, generally healthy except for the condition under study, willing and able to provide consent, adherent to complex protocols. The elderly were excluded. The very sick were too risky. Children were protected. The pregnant were forbidden. Racial and ethnic minorities were underrepresented; women were systematically excluded until the 1990s.

We had purchased internal validity at the cost of “external validity”: the ability to generalize findings beyond the narrow confines of the trial itself. We could prove a drug worked for the ghost, but we struggled to predict whether it would work for the grandmother with polypharmacy, the child with a rare variant, the patient whose ancestry and environment had shaped their biology in ways our trials never captured.

The fortress stood for decades. Until cracks began to appear.

III. The epistemology of the mean

To understand the depth of this problem, we must excavate the mathematical bedrock upon which the clinical trial rests. At its foundation lies the Central Limit Theorem,3 that miraculous promise that the mean of a sufficiently large sample will approximate the population mean. Francis Galton called regression to the mean “the most beautiful law in all statistics.” R.A. Fisher gave us the mathematics of randomization and the analysis of variance. These tools allowed us to extract signal from noise, to see the ghost emerge from the data.

But biology is not obligated to respect our statistical conveniences. In fact, like the greater universe, it is completely unaware of and indifferent to them. A normal distribution assumes that variation is random noise around a true central value: that patients deviate from the mean the way bullets scatter around a bullseye, their differences the product of measurement error and uncontrolled variance.

What if the “noise” is the signal? What if what we have been calling variation is actually heterogeneity: meaningful biological differences that determine whether a patient thrives or dies on a given therapy?

Let’s examine this in terms of music. In my 2002 medical school thesis on physiological harmonics (link above), I suggested that health is not a point on a number line but a key signature: a relationship among variables that together create concordance. Disease is not deviation from a mean; it is modulation into dissonance. A heart in normal sinus rhythm plays in predictable, consistent time; atrial fibrillation is the dissolution into arrhythmic chaos. In the composition Perpetual Drift, I traced how a simple heart block, first-degree and barely symptomatic, can degrade through second-degree Mobitz II into complete dissociation, then spiral into the polymorphic terror of torsades de pointes. Each transition is a key change, a harmonic shift.

The traditional RCT measures the average volume of the crash, but it is deaf to the melodic progression that preceded it. It aggregates a thousand different songs, squeezes them into a mean, and calls the result music.

IV. The molecular revolution and the epistemic crisis

Then came the genome. In 2000, the completion of the Human Genome Project promised a new epoch: personalized medicine, tailored to the individual’s genetic code. The reality proved more labyrinthine. We discovered that human genetic variation was both more extensive and more subtle than anticipated. Single nucleotide polymorphisms, single-letter typos in the three-billion-letter text of the genome, number in the millions. Some change everything; most do nothing; many interact in ways we still cannot predict. In truth, we know very little.

But in rare diseases, especially those of Mendelian origin, the promise crystallized. When a condition results from a single genetic error, a nonsense mutation in MECP2 causing Rett syndrome, a trinucleotide expansion in HTT causing Huntington’s disease, the old paradigm shatters. If there are only a few in the world with a particular mutation, how do you design a randomized trial? Who comprises the control arm? How many must die untreated to satisfy our epistemic demands?

The question is ethical vertigo: the recognition that our evidentiary standards, built to protect the many, have become obstacles to saving the particular. A barrier to the very essence of precision medicine.

V. A signal in the noise

In November 2025, the FDA published a brief sounding board in the New England Journal of Medicine. Co-authored by Commissioner Martin Makary and Vinay Prasad, it bore the understated title “FDA’s New Plausible Mechanism Pathway.” Within its three pages lay something remarkable, though its significance is easy to misread. This was not the arrival of a new era. It was something more modest and more important: a signal, a first step, a door cracked open onto a landscape we have only begun to map.

The catalyst was Baby K.J., a male neonate who presented within 48 hours of birth with carbamoyl-phosphate synthetase 1 (CPS1) deficiency. The enzyme defect prevented his body from processing dietary protein, causing ammonia to accumulate to neurotoxic levels. Without intervention, the natural history was clear: progressive brain damage with each hyperammonemic crisis, punctuated by hospitalizations, until the child grew large enough for liver transplantation, if he survived that long.

Genetic sequencing identified the specific biallelic mutations responsible. A team designed a bespoke therapy: lipid nanoparticles loaded with guide RNA and messenger RNA encoding an adenine base editor, targeted to repair that exact genetic error. The FDA processed a single-patient Investigational New Drug application in one week. Three consecutive doses were administered. The child’s ammonia levels normalized. He began tolerating dietary protein. The need for nitrogen-scavenger medications declined.

There was no control group. There were no randomized comparators. There was no placebo arm of doomed infants sacrificed to statistical rigor. The patient served as his own control, measured against the well-characterized natural history of untreated disease.

The pathway means that when a manufacturer demonstrates success across several consecutive patients with different bespoke therapies using the same platform technology, the FDA may grant marketing authorization not for each individual therapy, but for the platform itself: the base-editing system, the nanoparticle delivery vector, the cell therapy process. Subsequent therapies targeting different mutations can then be deployed under that umbrella approval.

Baby K.J. lived. That matters infinitely. But we must be careful not to mistake a beginning for an ending, a seed for a harvest. The pathway is not the transformation we need. It is evidence that transformation has become thinkable, that the regulatory architecture we inherited from the 20th century can evolve. What we do with that evidence will determine whether this moment becomes a turning point or a footnote.

VI. The work ahead

The pathway shows us what is possible, but it does not create what is necessary. Between the vision of precision medicine and its realization is an enormous expanse of unglamorous, essential work: the construction of infrastructure, the development of methods, the cultivation of relationships. None of this can be accomplished by fiat alone. All of it requires collaboration.

Natural history studies must be conducted with the same rigor we once demanded of control arms. If a person’s only comparator is the natural history of their condition, that history must be characterized with precision. Who will fund these studies? Who will ensure they capture the diversity of patients who will eventually receive treatment?

Post-marketing surveillance must be robust and durable. When a gene-editing therapy modifies a patient’s genome, the effects may unfold over years or decades. Off-target edits may have consequences that are invisible at first. Immune responses may evolve. This demands real-world data systems that persist long after the initial commercial push has faded, follow-up mechanisms that track patients across moves and insurance changes, and analytical methods capable of detecting true signals in noisy environments.

There is a temptation, when new regulatory flexibility appears, to view it as a concession extracted from reluctant gatekeepers. This framing is both historically inaccurate and strategically misguided. The pathway emerged not despite the FDA but because of it, because the agency recognized that existing frameworks could not accommodate what biology now makes possible.

Working shoulder to shoulder with regulators means something specific. It means investing in the infrastructure they need to evaluate novel evidence types. It means transparency about what we know and what we do not. It means participating in the hard work of defining standards and clinically validated endpoints, even when they constrain short-term commercial interests. It means treating post-marketing requirements not as burdens but as expressions of responsibility for the patients who receive our therapies in the real world.

VII. The N of 1 at scale

The phrase “precision medicine” has always contained a paradox. Medicine operates at scale: systems that treat millions, insurance schemes that pool risk across populations, research programs that aggregate data for statistical power. Precision implies the particular: this patient, this mutation, this moment. How do we reconcile mass with customization, the industrial with the artisanal?

Appropriately validated platform technologies can create precision at scale. A single base-editing system can address dozens of monogenic diseases. An antisense platform can target hundreds of pathogenic transcripts. An adoptive cell therapy framework with standardized manufacturing can be reprogrammed for different tumor antigens. An AI-enabled decision support system can recommend combination therapies calibrated to individual tumor genomics, adjusting its recommendations as each patient’s response unfolds. The marginal cost of designing the next therapy, once the platform is established, becomes manageable. Mass customization, the manufacturing paradigm that transformed industries from automobiles to consumer electronics, becomes applicable to medicine.

But manufacturing is the easy part. Harder is the epistemological challenge: generating evidence for therapies that may each be used in only a handful of patients, yet must collectively demonstrate safety and efficacy. This is the problem of the N of 1 at scale: proving that a methodology works across many single-patient applications.

The solution is not in more sophisticated trial designs conducted in isolation from clinical practice, but in dissolving the boundary between research and care altogether. When we embed these platform technologies into routine care across health systems, when every infusion center and community oncology practice and academic medical center operates on shared infrastructure, something remarkable becomes possible: we learn from each patient we treat. We intervene, we observe, we feed what we learn back into the system, and we improve. The next patient benefits from every patient who came before.

This is the vision of the learning health system applied to precision medicine. Each treated patient becomes not merely a recipient of care but a contributor to collective knowledge. The child with CPS1 deficiency whose ammonia normalizes after base editing teaches us something about editing efficiency, about immune response, about durability of effect. That knowledge, captured systematically, informs the treatment of the next child with a different urea cycle disorder. The oncology patient whose tumor responds unexpectedly to a novel combination reveals patterns that the AI system incorporates, refining its recommendations for future patients with similar molecular profiles.

The cycle must be continuous: intervene, observe, learn, improve, intervene again. Each iteration sharpens our understanding. Each patient encounter generates data that flows back into decision support systems, natural history registries, safety surveillance databases. The platform improves not through occasional version updates but through constant refinement, the way a recommendation algorithm learns from every interaction.

VIII. The convergence

We have been here before, or thought we were. The first generation of real-world data platforms promised transformation and delivered fragmentation: incompatible systems, inconsistent data standards, analytics that could not distinguish signal from noise. Electronic health records captured billing codes rather than biology. Registries accumulated data that no one analyzed. The infrastructure for learning health systems was built, but the learning never materialized.

What is different now is the convergence. Over a decade of investment in health information technology has created, despite its failures, a substrate upon which something better can be built. Genomic sequencing costs have collapsed a thousandfold. Artificial intelligence can capture structured data as an ambient scribe and extract patterns from datasets too vast and too complex for human comprehension. Regulatory science has evolved to accommodate novel evidence types. And the spirit of American innovation, that restless ingenuity that has always driven our most consequential advances, has turned its attention to the architecture of medical knowledge itself.

The pieces of the puzzle are coming together. Interoperable data standards are emerging. Cloud computing enables aggregation across institutions. Machine learning can control for confounding in observational data, detect safety signals in real time, identify patients likely to respond to specific therapies. The FDA has signaled willingness to accept evidence generated outside traditional trials. Payers are beginning to recognize that paying for data generation creates value. Patients, increasingly, demand to be partners rather than subjects.

Of course, the technical capability to build learning health systems does not ensure the political will to fund them, the institutional commitment to sustain them, or the social trust to make them acceptable. But the possibility that seemed utopian a decade ago now seems achievable. The question is no longer whether we can build systems capable of learning from the N of 1 at scale, but whether we will.

IX. The digital country doctor

What might come out of this work is a synthesis: a return to the philosophy of the Country Doctor, the obsession with the individual patient, the deployment of clinical judgment in the context of intimate knowledge, but equipped with tools s/he could never have imagined.

The 19th-century physician knew their patients across decades, accumulated pattern recognition through experience, and developed an intuitive sense for prognosis. But their knowledge was shallow vertically. They could palpate the spleen but not sequence the JAK2 mutation causing its enlargement. They could percuss the chest but not image the ground-glass opacities of pulmonary fibrosis or predict progression through algorithms trained on millions of CT scans. Their judgment was informed by dozens of patients; ours can be informed by millions.

The vision of 21st-century precision medicine inverts the Country Doctor’s limitations: vast mechanistic knowledge, genetic, transcriptomic, proteomic, metabolomic, coupled with computational tools to interpret it at the resolution of the individual. Each patient becomes not a deviation from the population mean but a unique trajectory through a high-dimensional state space, their biology a specific point that can be characterized, modeled, and targeted.

In oncology, this transformation is becoming visible. “Lung cancer” has decomposed into a taxonomy of molecular species: EGFR-mutant adenocarcinoma responds to osimertinib; ALK-rearranged disease to alectinib; tumors with high PD-L1 expression to pembrolizumab; KRAS G12C mutations to sotorasib. The average survival of “lung cancer patients” becomes a meaningless statistic; what matters is the natural history of this molecular subtype in this patient given this intervention.

X. Coda: the seed and the soil

James Lind summoned the statistical ghost aboard HMS Salisbury out of necessity. To prove that citrus prevented scurvy, he needed to demonstrate average efficacy. The ghost served us well for nearly three centuries. It gave us antibiotics and vaccines, chemotherapy and antiretrovirals and targeted therapies. It protected millions from ineffective or dangerous treatments. It established the principle that medical claims must be proven, not merely asserted.

But the ghost was always a compromise, a necessary simplification in an era when we could measure populations but not individuals, when we could characterize diseases by symptoms but not mechanisms, when we could statistically aggregate but not molecularly dissect.

We are beyond that era now. The genome is sequenced. The proteome is being mapped. Computational biology can model individual trajectories through state space. Gene editing can target single-nucleotide variants. The epistemological justification for the ghost has dissolved.

What remains is the ethical question: Having gained the capability to treat the particular, do we have the wisdom to wield it responsibly? Can we accelerate access without sacrificing safety? Can we embrace precision without encoding bias? Can we validate platforms without forcing dying patients to wait for statistical significance?

The regulatory signals we have received do not answer these questions. They ask them. They are seeds, planted in soil we have only begun to prepare. Whether they grow into the transformation precision medicine requires depends on what we do next: the infrastructure we build, the methods we develop, the partnerships we forge, the commitments we sustain.

The goal is not to force every patient into the same key signature, but to understand the particular song each person plays and to intervene not with the blunt instrument of population averages, but with the precision of a composer who knows that changing a single note, correcting a single base pair, can transform dissonance into resolution, silence into symphony, the inexorable trajectory toward death into the improbable grace of continued life.

We are learning, three centuries after Lind, that the choice was never between the aggregate and the particular, between rigor and compassion, between population health and individual salvation. The choice is whether we have the courage to begin the work that this moment demands: to labor alongside our peers and regulators to build systems capable of learning from the N of 1 at scale, to accept that the pathway we have been given is not a destination but a departure point.

Baby K.J. lived. That matters infinitely. But what matters more, in the long arc of medical history, is what we make of his survival: whether we treat it as a singular miracle or as the first note of a symphony we must now learn to compose, together.

The music has begun. Let us all rise to play our part.

Subscribe now

I. The Alchemists

Last week, the fog lifted from San Francisco, and for four days the city was bathed in a light so sharp it seemed to render everything visible. Those of us who gathered at the Westin St. Francis and across the satellite meetings of the JPM Healthcare Conference (JPMC) moved through this clarity with a steadiness that has been absent in recent years. The post-pandemic uncertainty had finally been metabolized. The zero-interest exuberance had burned off. Something more grounded had taken its place.

I walked the halls, attended the panels, sat through the presentations, and caught up with colleagues. And I found myself wondering not whether we could execute on our plans, but whether any of us fully appreciated what we were building.

The alchemists of our age no longer seek to transmute lead into gold. We seek something far stranger: to feed the disease of the body into a machine and receive, in return, its cure. To make biology computable. To augment the intuition that has driven every medical breakthrough in human history with the silent pattern-matching of neural networks, systems no less voracious than the most relentless human minds in their energy consumption.

Artificial intelligence was the promise echoing through every corridor of JPMC 2026. Eli Lilly and NVIDIA announced a billion-dollar partnership to build a co-innovation laboratory in South San Francisco. The ambition is to have robots conduct experiments in continuous loops, feeding results into models that propose new experiments, which the robots then execute, iteratively, until something works. The goal is not merely acceleration, but transformation: shifting discovery from hypothesis-driven intuition to computational search across vast chemical spaces. It is worth flagging this one for follow up.

Yet I found myself drawn not to these well-capitalized initiatives, but to the margins of the conference. To small teams building foundation models for biology from first principles because I think they understand something that is harder to see from inside large organizations: that artificial intelligence in drug discovery (and development) is not a tool to be bolted onto existing processes. It is a solvent that dissolves them. It demands a fundamentally different way of organizing the labor of understanding life, and the organizational physics of large companies do not always bend easily toward such dissolutions. This is not a criticism, but a structural observation about paradigm shifts. We have all seen how difficult it is to rebuild the ship while sailing it.

There is another class of company I have been watching, and saw again at JPMC with growing unease. These are firms that raised massive VC rounds or went public on the promise of AI-driven discovery, amassed vast data assets, and poured capital into the promise of automation, yet have not produced a single drug of meaningful clinical benefit. Platform stories are seductive precisely because they defer the question of whether the thing actually works. That deferral can last for years. Eventually, the question returns, and when it does, the answer is always binary.

II. The magic number

There is a number that lingered over the conference: $149.

This is the price, in dollars, that Eli Lilly proposed for a month’s supply of orforglipron, their oral obesity drug awaiting FDA approval. One hundred and forty-nine dollars. Less than a dinner for two in San Francisco. The cost of a streaming subscription, multiplied across a year.

The number landed with weight because we all understood what it signified. For the first time in the long history of human beings struggling with their own bodies, obesity is becoming a treatable condition at scale. Not through willpower alone. Not through the moral framework that has governed our understanding of weight for centuries. Through chemistry.

The GLP-1 agonists, Ozempic, Wegovy, Mounjaro, Zepbound, have already begun to reshape the landscape. But they have done so largely as scarce interventions, expensive and rationed, available to those with good insurance or deep pockets. What Lilly signaled was something different: the intention to manufacture at a scale and price point that could reach everyone who needs it.

The competitive dynamics are intensifying accordingly. Novo Nordisk launched the first oral Wegovy in early January, a week before JPMC opened, and their CEO Mike Doustdar spoke of 2026 as the year oral GLP-1s become accessible to the mass market. Lilly’s orforglipron, if approved, would follow shortly, with the added advantage of no dietary restrictions, a convenience that may prove decisive in a therapeutic class where adherence is everything. Behind them, the field is broadening: Pfizer, having spent roughly $10 billion to acquire Metsera after a bruising bidding war with Novo Nordisk, announced plans for ten Phase III trials in 2026, racing to launch by 2028. Albert Bourla was direct about what this means for a company that some say has struggled since the pandemic: “I don’t want to hear again that there are no catalysts for Pfizer.” Amgen’s MariTide, with its long-acting profile, adds yet another entrant. What was once a duopoly is becoming a crowded field.

On my last evening in San Francisco, waiting for my flight back to New York, I stopped at a restaurant near the gate. When my order arrived, it was far larger than I had expected, one of those portions that reminds you some restaurants operate on a different scale than the human stomach. The woman seated next to me glanced over and smiled. “You’re going to need a long jog after this,” she said.

“No,” I replied, “just a shot of GLP-1.”

She laughed. The man beside her leaned in. “Is this where we’re going?” he asked. “Stabbing ourselves so we can eat?”

It was a joke, but it wasn’t. In that brief exchange at an airport bar, I was the only one carrying the residue of JPMC but the three of us had a shared understanding: that a class of medications developed for diabetes had become, in the span of a few years, a shared cultural reference point for appetite, willpower, and the limits of self-control. The drugs had escaped the conference halls. They had entered the vernacular and become a punchline, which is how you know something has become real.

But sitting there, waiting to board, I found myself thinking not about the competition, or the pricing, or even the punchline. I was thinking about what lies on the other side of it. What happens when the body becomes correctable at scale? When appetite itself can be modulated?

The GLP-1s do not merely suppress hunger; they seem to quiet craving in some deeper register. Patients report diminished interest in alcohol, in gambling, in the compulsive loops that have always been part of human experience. Like so many other phenomena in biomedicine, the mechanism is not fully understood, but the pattern is consistent enough to suggest we are approaching something unique: a pharmacological switch for desire itself.

We had discussed this as a pricing story at JPMC, which of course it is. But it is also something more. We are on the threshold of a new relationship with the human body and we are crossing it in the language of market access and quarterly earnings and in jokes exchanged between strangers at airport gates. Perhaps that is the only way such thresholds get crossed: through commerce, through humor, through the slow normalization of what once seemed impossible.

I suspect we will look back on this moment and recognize that something profound was at stake, even if we lacked the vocabulary to name it at the time.

III. The new Silk Road

On the first day of the conference, AbbVie announced it had paid $650 million, with up to $4.95 billion in milestones to follow, for the rights to a molecule called RC148. It is a bispecific antibody, elegant in design, targeting both PD-1 and VEGF, a combination that may help the immune system fight tumors more effectively while overcoming resistance mechanisms that have limited earlier checkpoint inhibitors. The molecule was developed by RemeGen, in the Chinese city of Yantai.

AbbVie acquired the rights to sell it everywhere except China. The molecule will travel westward and the value will be realized in American and European markets.

This pattern repeated throughout the week. Roche signed deals with MediLink and Hansoh. Novartis partnered with SciNeuro and Zonsen PepLib. Pfizer’s own PD-1/VEGF bispecific came from a Chinese company. The modalities driving oncology’s cutting edge, antibody-drug conjugates, bispecific antibodies, radioligand therapies, are flowing into Western portfolios from laboratories in Shanghai and Suzhou and Hangzhou. Some of the most promising molecules of the next decade are being conceived in a country with which our political relationship grows more complicated by the year.

And yet, on those same stages, Johnson & Johnson’s Joaquin Duato announced $55 billion in American manufacturing investment. We are simultaneously building walls and extending bridges, investing in domestic infrastructure while deepening our dependence on foreign innovation. I thought of the Silk Road, that ancient network along which goods and ideas traveled in both directions, enriching and destabilizing every civilization it touched. We are building something similar now, with American infrastructure housing global innovation. Capital flowing east, molecules flowing west, and in the middle, a political conversation that has not yet caught up to the scientific reality.

In many ways, we are trying to protect our leadership without severing the arteries that feed it. The balance is delicate. The molecule RC148, born in Yantai, acquired in San Francisco, will soon be infused into the veins of patients with cancer who will never know or, some believe, never need to know where it came from.

This is the topology of the biomedical enterprise in 2026. Whether it can sustain the contradictions building within it is a question that hovered just outside the frame of conversations at JPMC.

IV. The cliff that isn’t

Patent cliffs, like taxes and death, are inevitable. They are the return of a truth that exclusivity allows us to defer: that the ownership of a molecule is temporary, that the profits flowing from a blockbuster drug are a lease, not a deed. The numbers underscored at the conference this year were sobering. More than $200 billion in annual revenue is at risk across the industry through the end of this decade. Merck’s Keytruda, Bristol-Myers Squibb’s Opdivo and Eliquis, AbbVie’s Humira, the drugs that have defined their companies for a decade are losing or will soon lose their protection. The challenge for these companies is navigating the terrain without catastrophic loss of altitude.

Merck’s Rob Davis announced a target of $70 billion in revenue from new growth drivers by the mid-2030s. To reach this number, Merck must essentially replace Keytruda, which accounted for nearly half the company’s revenue last year, with a portfolio of medicines still making their way through development. Davis characterized the coming loss of exclusivity as “more of a hill than a cliff.” This framing is, of course, understandable.

AbbVie offered a different case study, one positioned as being further along in its transition. Having already weathered Humira’s loss of exclusivity in 2023, the company has positioned Skyrizi and Rinvoq to reach $31 billion in combined sales by 2027, effectively replacing what Humira once provided. The execution has been impressive with the share capture faster than analysts expected. Bristol-Myers Squibb is targeting more than ten new medicine launches by 2030. The playbook is consistent across the industry: accelerate the pipeline, pursue strategic acquisitions, build the bridge to the other side before the ground shifts beneath us.

There is nothing cynical in observing that this is difficult. None of us are naive about the challenge as we have seen cliffs before. Some companies have navigated them; others have been diminished. I think what distinguished this year’s presentations was a quality of seriousness, a recognition that the margin for error has narrowed, that execution rather than aspiration will determine outcomes.

The next five years will make the difference clear.

V. The empty chairs

For the second consecutive year, the major health insurers did not attend.

UnitedHealth. CVS Health. Cigna. Humana. Elevance. The institutions that determine which treatments will be reimbursed, which interventions will reach patients, which innovations will translate from clinical trials into clinical practice, they were absent from the Westin St. Francis.

Their absence shaped the conference in ways that were felt by some and mentioned by a few.

We spent four days discussing the elegance of gene therapies, the precision of bispecific antibodies, the transformative potential of artificial intelligence, the coming democratization of obesity treatment. But the partners who will ultimately help determine how much of this reaches patients were elsewhere, engaged in conversations, that one can assume, we cannot be part of.

Needless to say, payers have their own pressures, their own constraints, their own logic. Those who create clinical value and those who adjudicate its economic viability are operating in increasingly distinct spheres. We speak of mechanisms and endpoints. They speak of utilization and total cost of care. The grammar is different and the concerns do not always align and, in many cases, the math bends in their favor.

Let’s consider the tensions embedded in some of the discussions around emerging therapies. A gene therapy that costs $2 million but may cure a disease in a single treatment. An obesity drug that could benefit a hundred million Americans but whose aggregate cost would reshape federal budgets. An Alzheimer’s treatment that might preserve cognition for a few additional months at a price that strains the definition of value. Who pays? On what terms? For how long?

These questions circled JPMC without resolution, because the people best positioned to answer them were not in the room. At some point, these conversations will have to converge. The longer the drift persists, the harder that convergence will become.

VI. After the conference

Stepping out of my last meeting into that improbable January light, I headed for the airport, the weight of the week measured by the stack of business cards in my pocket.

As I walked to my gate, scrolling through #jpm2026 on X, I saw a post claiming this had been the slowest JPM for news in years. In a narrow sense, that was true. There were no transformational acquisitions announced onstage. No single deal dominated the headlines. The mood was industrious rather than triumphant.

But I have learned to distrust the equation of news with significance. Some of the most important shifts announce themselves quietly, through the slow realignment of capital and conviction, through the moment when an industry stops promising transformation and begins the harder work of delivering it.

That is what I observed in San Francisco. An industry that has absorbed recent shocks and emerged more focused. GLP-1 companies building for scale rather than scarcity. AI partnerships moving from press releases to physical laboratories. Startups pursuing genuinely novel biology at the margins, largely unnoticed by the main stage but potentially more consequential than anything presented upon it. The quiet integration of Chinese innovation into Western portfolios. Patent cliffs enforcing a discipline that might otherwise have been deferred.

JPMC 2026 lacked the speculative electricity of earlier years. The sense that anything might be announced at any moment has given way to something more measured. Whether this reflects maturity or simply a different phase of the cycle is unclear. Perhaps the two are inseparable.

What I know is this: the fog has lifted. Many have seen themselves clearly and they have chosen execution over spectacle. What we build from here, whether the bridges hold, whether the science delivers, whether the economics align, will be the work of the next decade.

The conference ended. We returned to our responsibilities. The winter sun continued to illuminate the city, indifferent to what had been discussed within it.

Subscribe now

Cancer forces a particular kind of honesty. It resists simplification, punishes short time horizons, and exposes the assumptions embedded in the systems built around it. As the science has grown more precise, more molecular, and more data-rich, it has become increasingly clear that progress is shaped as much by institutional design as by discovery itself.

That reality framed my recent conversation with Dan McHugh, an investor whose career has unfolded at the seam between biology and capital. His path, from Stanford bioengineering and the Greenleaf Lab, through Bain & Company and Emerson Collective, to the co-founding of Tune Therapeutics, has repeatedly placed him in moments where scientific promise encounters structural constraint. Across those settings, a consistent pattern emerges: progress in cancer is governed not only by what we know, but by how patiently and coherently institutions are willing to act on incomplete knowledge.

Dan’s time at Emerson Collective, founded by Laurene Powell Jobs, sharpened this view. The organization’s model, lending philanthropy, investment, and long-horizon advocacy, treats time as an asset rather than a liability. Working within that structure clarified something important: capital architecture is not neutral. How money is organized determines which kinds of science are allowed to persist long enough to matter.

That insight now animates Yosemite, the cancer-focused investment platform founded by Reed Jobs, where Dan serves as an investor. Yosemite was built around a specific ambition: to alter the lethality of cancer over a human lifetime. The team approaches this mandate as a design challenge, aligning capital, incentives, and time horizons with the biological realities of the disease, including long development cycles and ambiguous early signals.

Much of our discussion centered on early-stage cancer science, where those realities are most pronounced. Foundational work often precedes validated endpoints, clean hypotheses, or commercial readiness. Traditional venture structures can struggle here, not because the biology lacks importance, but because surrounding systems are optimized for speed and early legibility. Yosemite's hybrid model, pairing venture investment with grant-based risk-taking, reflects a deliberate strategy to preserve optionality, allowing exploration to mature without forcing premature narratives or artificial milestones.

Dan’s experience translating epigenetic science at Tune Therapeutics deepened this perspective. There, he saw firsthand how reimbursement logic, regulatory sequencing, and market expectations quietly shape what is pursued, how long it is allowed to develop, and which uncertainties are tolerated. These forces do not merely follow science, they actively bend its trajectory.

What makes Yosemite distinctive is not just its structure but the trust underlying it. Dan’s long-standing friendship with Reed provides continuity of purpose that supports long-horizon decision-making. In cancer investment, where feedback loops are slow and ambiguity is unavoidable, that kind of alignment functions as differentiating infrastructure, shaping how uncertainty is absorbed and how conviction is sustained when outcomes are not immediately legible.

Throughout our conversation, reimbursement, regulation, and market structure surfaced as governing forces in biomedicine. They encode assumptions about time, value, and acceptable uncertainty, amplifying some paths while quietly filtering out others. As scientific capabilities accelerate, the distance between what biology allows and what institutions can absorb becomes increasingly consequential.

The implication is a shift in emphasis. Progress in cancer depends not only on better tools or deeper datasets, but on whether our investment and institutional architectures can carry complexity over time. Biology will continue to adapt and resist. The determining factor will be whether the systems surrounding it evolve with comparable agility and whether those shaping capital and incentives treat that work as the systems-design challenge it is.

Subscribe now

There is a distinct difference between managing an organization and transforming one. In the latest episode of Precision Signals, I sat down with someone who has built a career on the latter: Dr. Karen Knudsen, the CEO of the Parker Institute for Cancer Immunotherapy (PICI).

Karen describes her upbringing as that of a “nomad” and an “army brat,” living in 11 different houses before graduating high school. That early exposure to constant change, and the stark contrast she witnessed between the secure healthcare of the military and the fragility of civilian access, instilled a certain fearlessness in her approach to leadership. Her career trajectory has been a climb through every altitude of the oncology ecosystem: from the molecular level as a yeast geneticist, to the clinical level running a massive service line at Jefferson Health, to the national stage leading the American Cancer Society (ACS), and now, to the intersection of philanthropy and venture capital at PICI.

Our conversation covered the unique mechanics of the PICI model, which effectively operates as venture philanthropy on steroids. We often talk about the “Valley of Death” in drug development, where promising science dies before it can attract commercial capital. PICI is attacking this problem by refusing to just be a check-writer^. Instead, they operate as a non-profit that holds the keys to intellectual property, identifying high-risk, high-reward science and then “giving birth” to for-profit companies to bring those therapies to the clinic.

By maintaining visibility into the IP across multiple rival institutions, PICI can connect dots that others miss, combining a discovery from one lab with a vector from another to create a viable company^. Karen highlighted recent successes like Dispatch Bio examples of this active de-risking strategy. It challenges the traditional boundaries between philanthropy and venture capital, proving that non-profits can be aggressive players in market-driven solutions.

Karen admits she is not a “keep the trains running” CEO. She is there to build new tracks and her approach offers a blueprint for how to navigate complex ecosystems: focus on the mission, demand business rigor, and don’t be afraid to dismantle the silos that slow us down.

Subscribe now

In the late nineteenth century, America’s railroad barons believed they understood the future with a confidence born of decades of industrial success. They had constructed an infrastructure meticulously sized for the movement of industrial freight—grain harvests from the Midwest, iron ore from Minnesota, petroleum from Pennsylvania—and they operated under the reasonable assumption that if they continued to expand track mileage and increase capacity along established corridors, economic demand would follow the familiar patterns that had governed growth since the Civil War. Their managers had achieved something approaching perfection in the machinery of precision scheduling and bulk logistics, orchestrating the movement of tonnage across continental distances with a reliability that represented the apex of nineteenth-century organizational achievement. The system functioned with remarkable efficiency until the fundamental character of the underlying economy began to transform in ways that quickly made their architectural assumptions obsolete.

By the 1890s, the emergence of consumer markets had begun to fundamentally reshape the flow of goods across the American landscape. The predictable rhythm of long-haul industrial shipments, which had defined railroad economics for a generation, gave way to an accelerating volume of small, irregular freight movements that responded to consumer demand rather than seasonal agricultural cycles or industrial production schedules. Railroad systems that had been engineered with exquisite precision for moving heavy loads along fixed routes found themselves increasingly unable to accommodate the operational complexity inherent in short-term hauls, high-fragmentation traffic patterns, and decentralized demand signals that could not be anticipated months in advance. The managers of the Pennsylvania Railroad, once celebrated throughout the business world for their extraordinary ability to choreograph large freight movements across their network, discovered themselves increasingly overwhelmed by the sheer operational complexity of coordinating thousands of small shipments with heterogeneous origins, destinations, and timing requirements. The system began to falter not because leadership had become incompetent or complacent, but because an architecture that had been brilliantly optimized for one economic era could no longer support the fundamentally different demands that characterized the next phase of industrial evolution.

The lesson embedded in this historical moment is rather straightforward even if it proved institutionally difficult to absorb: progress in the external environment eventually overwhelms legacy infrastructure regardless of how polished, professionally maintained, or conceptually elegant that infrastructure may be. The railroads did not fail because they were poorly managed or inadequately capitalized. They encountered profound difficulties because they had been built for a world that no longer existed and the gap between their architectural capabilities and the requirements of the emerging economy widened faster than incremental adaptation could bridge.

This form of structural mismatch (the growing incompatibility between what modern science demands and what existing systems were designed to deliver) defines the predicament in which contemporary healthcare and biomedical research find themselves at this moment in their institutional evolution.

The architecture of an earlier era

Modern biomedical infrastructure was designed around a constellation of conceptual and technological assumptions that accurately reflected the realities and constraints of late-twentieth-century science. Clinical trials assumed episodic patient encounters organized around scheduled clinic visits, low-frequency safety monitoring that corresponded to the temporal resolution of available measurement technologies, modest data volumes that could be recorded on paper forms and later entered into relatively simple databases, and unidimensional readouts such as RECIST criteria for tumor assessment, ejection fraction measurements for cardiac function, or serum biomarker concentrations that could be captured as single numeric values. Healthcare delivery systems assumed site-centric care teams that operated within the physical boundaries of hospitals and clinics, sparse laboratory and imaging data generated at discrete timepoints separated by weeks or months, structured billing codes as the dominant data vocabulary through which clinical activity could be captured and communicated, and workflows that were fundamentally mediated by human documentation rather than continuous automated data capture. Regulatory science developed around trial designs predicated on prespecified treatment arms with fixed definitions, scheduled assessment intervals that reflected both the biology of disease progression and the practical constraints of patient monitoring, narrow primary endpoints that could be measured reliably with available tools, and controlled case-report forms that channeled clinical observations into standardized categories amenable to statistical analysis.

All of these frameworks emerged through a logical process of adaptation to the scientific constraints and technological capabilities of their time. They created intellectual and operational order in what was necessarily a low-resolution observational environment and they produced decades of genuine if incremental progress in understanding disease biology and developing therapeutic interventions. Like the polished scheduling machinery that enabled the Pennsylvania Railroad to move freight with unprecedented reliability, these systems were genuinely efficient and effective for the specific challenges they were constructed to solve. The difficulty is that scientific and technological progress has fundamentally outgrown the architectural assumptions on which they rest, creating a widening gap between capability and infrastructure that now threatens to become the binding constraint on translational progress.

Science has shifted; infrastructure has not

The dominant pattern of contemporary biomedicine is a fundamental shift toward fine-grained, multimodal, continuous measurement of human biological state across temporal and spatial scales that were simply inaccessible a generation ago. The field has evolved from macroscopic, episodic observations captured at discrete clinical encounters to a dense, high-dimensional data universe that encompasses multiplexed proteomic and transcriptomic signatures capable of resolving thousands of molecular species simultaneously, radiomic and volumetric imaging analyses that extract quantitative features from medical images at scales ranging from millimeters to microns, circulating tumor DNA trajectories that track tumor burden and clonal evolution through serial blood sampling, single-cell atlases that map cellular heterogeneity within tissues at unprecedented resolution, high-frequency physiologic monitoring from wearable sensors and ambient monitoring systems that capture continuous streams of cardiovascular and metabolic data, natural-language patient narratives that document symptoms and experiences in their own vocabulary, and algorithmically derived features extracted through machine learning methods applied to imaging, digital pathology, and clinical text that identify patterns invisible to direct human observation.

Each individual signal represents a relatively modest increment to the observational base—just as individual consumer shipments were small additions to railroad freight volume—but the aggregate complexity of integrating these heterogeneous data streams, reconciling their different temporal resolutions and semantic structures, and extracting coherent biological meaning from their interactions has grown explosively. Biological state is now routinely observed across more dimensions, at higher temporal resolution, and through more heterogeneous measurement modalities than at any previous moment in the history of experimental medicine.

The challenge is that the institutional systems responsible for generating clinical evidence and delivering patient care continue to operate as though the primary unit of medical data remains a manually recorded laboratory value or a radiologist’s categorical judgment rendered at six-week intervals according to predefined assessment schedules. This represents far more than a temporary inefficiency or a lag in institutional adaptation. It constitutes a fundamental architectural incompatibility between the measurement capabilities that modern science has developed and the data infrastructure through which those measurements must flow to generate clinical evidence and inform therapeutic decisions.

Operational strain as a symptom of structural mismatch

The stress fractures now visible across the healthcare and clinical research ecosystem bear a striking resemblance to the operational breakdown experienced by railroad systems attempting to accommodate freight patterns for which they were never designed. The manifestations of this structural mismatch appear throughout the evidence generation pipeline, and while they are often interpreted as isolated operational failures or process deficiencies, they are more accurately understood as symptoms of a deeper architectural problem that cannot be resolved through incremental process improvements.

Data heterogeneity now exceeds the absorptive capacity of existing information systems by margins that continue to widen. Electronic health record platforms cannot natively ingest or harmonize multimodal data streams because they were architected around structured fields optimized for billing and clinical documentation rather than high-dimensional scientific measurement. Clinical trial platforms cannot reconcile continuous monitoring data with assessment schedules predefined in protocols written months or years before study initiation. Regulatory submissions struggle to integrate imaging-derived endpoints, multimodal profiling, and real-world evidence streams into coherent evidentiary packages because the regulatory framework was constructed around the assumption that evidence would arrive in discrete, prespecified formats with clear temporal boundaries between data generation and analysis.

Manual adjudication processes are being asked to scale beyond the fundamental cognitive limits of human information processing. RECIST assessments of tumor burden, adverse event classification according to standardized toxicity criteria, eligibility screening against increasingly complex inclusion and exclusion criteria, chart abstraction to populate case report forms, and safety review of accumulating clinical data all continue to rely overwhelmingly on human judgment applied to individual cases. The inter-rater variance in these judgments is measurable and substantial. The cost of maintaining these manual processes is escalating as data volumes grow. The throughput is fundamentally capped by human cognitive capacity and available time, creating bottlenecks that no amount of additional staffing can fully resolve because the problem is architectural rather than simply a matter of insufficient resources.

Trial conduct has become progressively slower, more brittle, and more expensive in ways that reflect structural constraints rather than operational inefficiency. Patient accrual suffers because eligibility models remain anchored in an earlier era’s data reality, requiring manual verification of criteria that could in principle be assessed computationally if the data infrastructure supported such assessment. Monitoring is constrained by episodic clinic visits because continuous remote monitoring has not been integrated into core trial operations and regulatory expectations. Data cleaning and query resolution can absorb more calendar time than scientific analysis because data flows through manual transcription and entry processes that introduce errors requiring labor-intensive correction. These failure modes are the predictable consequences of attempting to execute modern science using infrastructure designed for a lower-dimensional, lower-velocity observational environment.

Artificial intelligence innovations encounter persistent obstacles not in model development but in clinical validation, where they collide repeatedly with the inconsistent ground truth generated by legacy measurement systems. When reference labels for imaging interpretation or pathological classification vary substantially across institutions, across individual human raters, and even across serial assessments by the same rater, modern machine learning methods cannot produce stable, generalizable models regardless of algorithmic sophistication or training data volume. The bottleneck preventing AI from realizing its potential in clinical applications is not computational capacity or algorithmic innovation. It is the absence of reliable ground-truth infrastructure against which models can be clinically validated and their performance measured with sufficient precision to support clinical deployment and patient care.

These phenomena are not isolated operational failures that can be addressed through targeted interventions. Taken together, they constitute a strong systemic signal indicating that existing infrastructure has become fundamentally misaligned with the dimensionality, velocity, and granularity that characterize modern biomedical science. Today, the gap between measurement capability and infrastructure capacity is no longer stable. It is also widening and the consequences are increasingly visible in the form of delayed translations, failed trials, and scientific capabilities that cannot be effectively deployed in clinical settings.

The illusion of precision in legacy systems

Railroad managers took justifiable pride in the high-polish scheduling machinery they had perfected over decades of operational refinement. Similarly, the contemporary clinical research ecosystem maintains considerable pride in its process discipline—standard operating procedures that specify every aspect of trial conduct, monitoring plans that detail oversight activities with bureaucratic precision, case report forms that channel observations into predefined categories, adjudication committees that apply standardized criteria to ambiguous cases, eligibility criteria that attempt to define patient populations with contractual specificity, and imaging guidelines that prescribe technical parameters and interpretation standards. Each procedural layer reflects a deeply embedded institutional belief that precision in scientific measurement emerges primarily from procedural constraint and standardization, and that reproducibility can be achieved by progressively tightening the specifications that govern how observations are made and recorded.

The difficulty is that these structures produce a particular form of narrow precision—reproducibility in the application of predefined measurement protocols—rather than high-dimensional accuracy in capturing the actual biological phenomena under investigation. They optimize the consistency with which an outdated measurement system is applied rather than the fidelity with which that system captures the underlying scientific signal. A RECIST assessment can be performed with high inter-rater reliability while still providing a fundamentally impoverished representation of tumor dynamics compared to what volumetric imaging and radiomics can reveal. An adverse event can be classified consistently according to CTCAE criteria while failing to capture the temporal trajectory, severity fluctuations, and mechanistic relationships that continuous monitoring would illuminate. An eligibility determination can be executed according to protocol specifications while excluding patients who would benefit from treatment because the eligibility criteria were written before biomarkers became available that could identify relevant biological subgroups.

The more aggressively the system tightens these procedural controls—mandating additional documentation, expanding monitoring checklists, demanding stricter adherence to predefined interpretive frameworks—the more clearly it reveals the underlying architectural mismatch. No amount of process discipline can compensate for the fundamental reality that the contemporary data ecosystem requires automated integration of heterogeneous data streams, continuous monitoring across multiple biological scales, and high-resolution phenotyping that manual systems cannot deliver regardless of how carefully they are executed.

The legacy infrastructure is not failing because it is poorly operated or inadequately resourced. It is failing because it represents over-polished machinery that was brilliantly optimized for the wrong world, and additional polish only makes the mismatch more apparent.

A new evidence architecture

If the railroad systems of the 1890s had attempted to survive through incremental process improvements—more detailed scheduling procedures, additional clerks to track small shipments, expanded documentation requirements—they would have collapsed entirely under the operational burden. Instead, those that survived underwent fundamental adaptation through the development of new freight-handling systems capable of processing heterogeneous cargo efficiently, new scheduling algorithms that could accommodate irregular demand patterns, and new commercial models that aligned pricing with the economic realities of small-shipment freight. The transformation was neither rapid nor easy and many railroads failed before it was complete, but those that succeeded did so by rebuilding their operational architecture rather than perfecting their existing procedures.

Biomedicine requires a comparable transformation—not cosmetic modernization or incremental digitization, but a foundational redesign of how clinical evidence is generated, validated, and interpreted. Several architectural elements are essential to this transformation and, while they can be described separately, they function as an integrated system in which each component depends on and enables the others.

High-resolution, multimodal data must transition from exceptional additions to clinical trials to become the default input for evidence generation. Trials and real-world evidence platforms must be capable of ingesting imaging-derived measurements, molecular profiling results, physiologic monitoring streams, and narrative patient reports not as supplementary attachments to core data collection but as first-class, semantically integrated inputs that participate fully in endpoint definition, statistical analysis, and clinical interpretation. This requires fundamental changes in data architecture, moving from document-centric storage systems (both physical and digital documents) to graph-based semantic models that can represent complex relationships among heterogeneous observations and support queries that span multiple data modalities and temporal scales.

Autonomous extraction of ground truth through validated AI systems must progressively replace manual measurement and adjudication for tasks where algorithmic methods can demonstrate superior reliability, consistency, and scalability. This transition depends critically on the availability of robust reference standards against which algorithmic performance can be rigorously validated—volumetric imaging measurements traceable to physical phantoms, digital pathology assessments benchmarked against quantitative molecular markers, structured safety signals extracted from clinical narratives and validated against quantifiable patient outcomes. The goal is not to necessarily eliminate human judgment but to deploy it more effectively by reserving it for genuinely ambiguous cases while automating the extraction of measurements that can be made reliably by computational methods.

Continuous endpoints must replace episodic assessments as the primary framework for measuring therapeutic effects and disease trajectories. Tumors do not grow in discrete jumps synchronized to clinic visit schedules. Toxicities do not emerge instantaneously at their maximum severity. Physiologic states evolve continuously along trajectories that contain far more information than snapshots captured at arbitrary timepoints. Evidence systems must be rebuilt to model rates of change, transition dynamics between disease states, and temporal relationships between interventions and biological responses rather than reducing these complex phenomena to categorical outcomes assessed at fixed intervals. This requires both new statistical methods capable of analyzing continuous trajectories and new data infrastructure capable of capturing high-frequency measurements across the relevant biological scales.

Integrated clinical study and real-world data infrastructure must emerge to replace the artificial separation currently maintained between clinical trials, electronic health records, disease registries, and patient-generated data streams. This separation reflects historical contingencies and institutional boundaries rather than any fundamental scientific or logical necessity. Future evidence models will necessarily unify these data streams within a shared semantic architecture that can track individual patients across care settings and time, link trial participation to long-term outcomes observed in routine care, and enable evidence generation that draws on the full spectrum of available observations rather than only those captured within the formal boundaries of controlled trials. Building this infrastructure requires solving formidable technical challenges in federated data access, privacy-preserving computation, and semantic harmonization across heterogeneous data sources, but these are tractable engineering problems rather than fundamental scientific barriers.

Regulatory pathways must reorient themselves to center on validation of measurement reliability rather than adherence to prespecified data formats. The central regulatory question for new measurement modalities and analytical methods should evolve from “Was this approach prespecified in the protocol before data were examined?” to “Has this measurement been validated to demonstrate clinically acceptable reliability, and is the evidence of that validation sufficiently rigorous to support regulatory decisions?” This reorientation would align regulatory expectations with the realities of algorithmic measurement and multimodal evidence while maintaining the fundamental regulatory commitment to ensuring that therapeutic decisions rest on reliable measurements of clinical benefit and harm. It requires regulatory science to develop new frameworks for evaluating algorithmic systems and to build institutional capacity for assessing validation evidence across diverse measurement modalities, but it does not require abandoning the core principles that have made regulatory oversight effective.

These shifts are neither optional enhancements to current practice nor distant aspirations for some hypothetical future state. They are architectural prerequisites for translating the measurement capabilities that modern science has already developed into clinical impact that improves patient outcomes. The scientific foundation for this transformation already exists. What remains is the institutional and engineering work of building the infrastructure that can support it.

The rails of biomedicine must be rebuilt

Science has undergone a fundamental transformation in its ability to measure human biology across scales and modalities that were inaccessible when our current infrastructure was designed. The institutional systems built to generate clinical evidence, evaluate therapeutic interventions, and guide clinical practice have not undergone a commensurate transformation. The consequence is a widening gap between what we can measure about disease biology and therapeutic response and what our translational infrastructure can effectively act upon to improve patient outcomes.

Closing this gap requires rebuilding the evidence architecture at a fundamental level—not through the addition of more procedural controls around outdated measurement systems, but through realignment of infrastructure with the full dimensionality, temporal resolution, and semantic complexity that characterize modern biological science. This transformation will require sustained investment, institutional courage to abandon deeply embedded practices that no longer serve their intended purpose, and the development of new regulatory frameworks that can ensure reliability without constraining innovation. The technical challenges are substantial but tractable. The institutional challenges are more formidable, requiring coordination across organizations with different incentives and cultures that have proven resistant to change.

Until this realignment occurs, clinical research and patient care will continue to exhibit the characteristic dysfunction of the 1890s railroads: institutional structures that remain elegant in their formal organization, professionals who take justifiable pride in the precision with which they execute established procedures, and a progressive inability to carry the scientific cargo that represents the future of medicine.

The question is not whether this infrastructure will eventually transform. Economic and scientific pressures will ultimately force adaptation. The question is whether that transformation will be led deliberately by institutions that recognize the structural nature of the challenge they face, or whether it will occur through the slower, more painful process of institutional failure and replacement. The choice between these paths remains open, but the window for deliberate action is narrowing.

Subscribe now

Progress in cancer research depends on three conditions operating simultaneously: consolidating established knowledge, integrating emerging technologies, and preparing for capabilities that don’t quite exist yet. This requires clarity about which established priorities deserve sustained investment and which “gold standards” deserve retirement (technically, there are no gold standards), which emerging areas are reshaping the field right now, and which future directions could fundamentally transform cancer biology and therapy. I believe progress emerges not from choosing between proven principles and novel capabilities, but from orchestrating them together in ways that expand what is scientifically and clinically achievable.

Prelude: the three-horizon problem

The history of cancer research show us that breakthroughs often come from the patient accumulation of foundational knowledge meeting new technological capability. The identification of oncogenes in the 1970s required decades of virology and molecular biology. Checkpoint immunotherapy, recognized with the 2018 Nobel Prize, built upon a century of tumor immunology before technological advances made clinical translation possible.

Science, it seems, rewards patience more generously than prophecy.

Today, we find ourselves conducting three experiments simultaneously. The first encompasses established areas where the scientific questions are clear and the infrastructure robust—these require continued investment to realize their full potential. The second includes emerging technologies already disrupting traditional approaches, reshaping how we conceptualize and investigate cancer. The third encompasses capabilities that have yet to penetrate cancer research but carry substantial potential for disruption.

The challenge is not choosing between these timelines, but orchestrating them into a coherent research ecosystem. As is often the case in science, the difficulty is not in having too few ideas, but in having too many simultaneously, each demanding resources and attention in different measures.

Movement I - Andante sostenuto: the non-negotiable core

On maintaining momentum in established territories

The persistence of precision

The initial promise of precision oncology and targeted therapeutics —matching molecular aberrations to targeted agents—has proven more complex than anticipated, as biology always tends to be when interrogated closely. Response rates to targeted therapies often plateau even in biomarker-selected populations with resistance an inevitable reality in nearly all cases, a humbling reality.

I think the next phase of the journey about integration: connecting molecular profiling with clinical decision-making in real-time, incorporating pharmacokinetic variability, and accounting for intra-tumoral heterogeneity that single-biopsy approaches cannot capture. The components of this system already exist: molecular testing platforms, computational pipelines, and clinical frameworks capable of embedding data into care. What does not yet exist is the system itself—a unified infrastructure where these elements operate as one continuum. The scientific challenge, therefore, is not invention but orchestration: moving from correlation (this mutation associates with response) to causation (this molecular circuit determines response), and from static classification to dynamic prediction. We have built the instruments; what remains is learning how to conduct them in harmony—and deciding where, precisely, to aim the telescope.

The immune-tumor dialectic

Tumor immunology and immune evasion mechanisms have transitioned from theoretical curiosity to clinical imperative. Yet current immunotherapies benefit only subsets of patients, and we’re still unable to predict with precision who will respond.

The tumor-immune relationship is not binary as it may seem—immunogenic versus non-immunogenic—but a continuously negotiated détente, with tumors actively shaping their immune environment while immune cells exert selective pressure driving tumor evolution.

Understanding this co-evolutionary dynamic requires moving beyond static snapshots toward longitudinal studies that capture tumor–immune interactions across treatment and disease progression. Biomarkers remain essential—but their real value lies in clinical validation, not mere discovery. That validation demands precompetitive collaboration and data sharing on a scale beyond the reach of any single institution or company. Only through such collective frameworks can we translate mechanistic insight into predictive, clinically reliable markers of response and resistance. I believe the goal, then, is not to multiply candidate biomarkers but to establish those that genuinely alter therapeutic decisions and outcomes as true surrogates—an endeavor as scientific as it is infrastructural.

The resistance imperative

Therapeutic resistance is the central clinical challenge in modern cancer therapies. Most anticancer therapies eventually fail, not because the initial target was wrong but because biological systems adapt. Under selective pressure, cancer cells exploit redundancy and plasticity within signaling networks, adapting through genetic and epigenetic reprogramming.

Recent studies reveal that resistance is frequently polyclonal and spatially heterogeneous: distinct tumor regions, and even subclones within them, adopt different survival strategies ranging from pathway reactivation to phenotypic state shifts and microenvironmental remodeling. Capturing these dynamics requires experimental frameworks that integrate spatial, temporal, and multi-omic analyses, coupled with in vivo and real-world data to trace clonal evolution under therapeutic pressure.

Meeting this challenge demands anticipatory rather than reactive strategies—therapies designed to pre-empt adaptation by targeting evolutionary constraints, synthetic lethal interactions, or resistance-prone cellular states. The alternative—serial deployment of single agents until resistance reemerges—is not precision medicine but sequential attrition, and its eventual outcome is predictably defeat.

Genomic infrastructure as enabling layer

Cancer genomics and molecular profiling define the empirical foundation of precision medicine. Large-scale sequencing efforts have catalogued millions of somatic mutations across cancer types, creating databases of impressive size and at times overwhelming complexity. The functional significance of most variants, however, remains unclear—we have compiled an extensive dictionary of a language we only partially understand.

I believe the next phase requires moving from descriptive cataloging to functional annotation: which alterations are drivers versus passengers (and are these designation even valid today), how mutations interact epistatically, and how genomic context influences therapeutic vulnerability. Importantly, genomics alone is insufficient. Integration with transcriptomic, epigenomic, proteomic, and metabolomic data—true multi-omic characterization—is necessary to capture the full complexity of tumor biology. The tools exists; the analytical frameworks for integration are still maturing. We have gathered the ingredients with the recipe still in draft form.

Movement II - Allegro con brio: technologies reshaping the present

On emerging capabilities disrupting traditional approaches

Computational intelligence enters the laboratory

Artificial intelligence and machine learning are at the cusp of transforming how we do cancer research. AI’s impact operates across scales: from classifying histopathological images with superhuman accuracy to integrating multimodal datasets that exceed human cognitive capacity for synthesis by several orders of magnitude.

The most consequential shift is conceptual. Traditional cancer classification relies on organ of origin and histological appearance—a system dating to the 19th century and showing its age. AI enables classification by mechanism: tumors grouped not by where they arose but by the molecular circuits driving their behavior. This mechanistic understanding facilitates treatment selection irrespective of anatomical site, exemplified by tissue-agnostic approvals for targeted therapies.

Location, it turns out, is not destiny.

Beyond classification, machine learning enables causal inference from observational data, computationally predicting regulatory networks and signaling dependencies without exhaustive experimental validation of every node and edge. Real-time treatment response assessment using continuous data streams—imaging, liquid biopsies, wearable sensors—becomes feasible with AI-powered integration. The bottleneck is no longer data generation but data interpretation and validation, and here computational approaches are not merely helpful but essential. We have moved from data scarcity to data abundance, which presents its own distinctive pathology.

Seeing tumors in context

Spatial multi-omics and tissue architecture mapping address a fundamental limitation of traditional molecular profiling: the loss of spatial information. Bulk sequencing homogenizes tumor and stromal cells; single-cell approaches provide cellular resolution but sacrifice spatial relationships. This is rather like analyzing a city by grinding it into powder and measuring the average properties—technically informative, but missing something essential about how the city actually functions.

New technologies—spatial transcriptomics, multiplexed imaging, and high-parameter mass cytometry—preserve tissue architecture while providing molecular detail. The implications extend beyond methodological improvement. Tumors are not homogeneous masses but spatially organized ecosystems with geography, neighborhoods, and borders. Proximity matters: cancer cells at the invasive margin face different selective pressures than those in the core; immune cells adjacent to tumor nests versus distant stroma differ in functional state and therapeutic relevance. Therapeutic gradients—drug penetration decreasing with distance from vasculature—create spatially variable selection pressures driving heterogeneous resistance.

Spatial profiling enables multi-scale systems modeling, connecting molecular signaling within individual cells to cellular interactions within tissue neighborhoods to organ-level tumor architecture. This hierarchy—molecules to cells to tissues to organisms—must be captured to understand cancer as an emergent phenomenon. Reductionism has served us well; it is time to reassemble what we have methodically taken apart.

Interrogating tumors without surgery

Liquid biopsy technologies and circulating biomarkers offer something genuinely useful: tumor profiling without invasive procedures. Circulating tumor DNA (ctDNA), tumor cells, extracellular vesicles, and proteins shed into blood provide molecular information about tumor state. The advantages are multiple: serial sampling to capture temporal evolution, minimally invasive procedures enabling frequent monitoring, and potential capture of tumor heterogeneity better than single-site biopsies.

Clinical applications, however, remain largely aspirational rather than routine. These technologies offer extraordinary technical potential: real-time, minimally invasive insights into tumor dynamics across space and time. Yet translation into broad clinical use has been constrained by a familiar bottleneck—clinical validation. Demonstrating that these assays meaningfully improve outcomes requires longitudinal, multi-institutional datasets of a scale no single organization or company can assemble alone.

The next phase, therefore, is not further technological refinement alone but collaborative standardization: defining which analytes, which assays, and which clinical contexts merit deployment. Without precompetitive frameworks to share data and harmonize evidence, the field risks remaining technologically advanced but clinically underpowered. We already possess instruments capable of reading tumors’ correspondence; what’s missing is the collective infrastructure to interpret that correspondence with confidence and consequence.

Resolution to see individual actors

Single-cell and single-nucleus profiling resolve cancer’s fundamental heterogeneity by examining tumors one cell at a time. Tumors comprise diverse cancer cell populations, immune cells, stromal cells, and vascular cells, each with distinct molecular states. Bulk profiling averages across this diversity, obscuring critical biology—a methodological choice that has perhaps contributed more to experimental convenience than biological insight.

Single-cell technologies reveal cancer cells as occupying a continuum of states rather than discrete categories. Cell plasticity—the ability to transition between states in response to microenvironmental signals or therapeutic pressure—is a defining feature. These state transitions can confer resistance: epithelial cells adopting mesenchymal features to survive targeted therapy, differentiated cells reverting to stem-like states to escape immune surveillance.

Cancer cells, it seems, are remarkably versatile when their survival depends upon it.

Moreover, tumors often evolve clonally. Initial driver mutations expand, but subsequent mutations create subclones with distinct properties. Single-cell sequencing tracks this clonal architecture and its evolution under treatment, revealing tumors as ecosystems undergoing Darwinian selection in real-time. The therapeutic implication is sobering: treatments that eliminate the dominant clone may simply accelerate expansion of resistant subclones already present at low frequency.

We are playing chess against an opponent who makes multiple moves simultaneously and never forgets a losing position.

Movement III - Scherzo: Capabilities on the horizon

On future directions not yet emergent in cancer science

Engineering biology as therapy

Synthetic biology and programmable cellular systems extend beyond current CAR-T cell therapy toward truly engineered living therapeutics. The vision encompasses cells programmed with genetic circuits that sense tumor microenvironmental signals—hypoxia, specific cytokines, tumor-associated antigens—and respond with calibrated therapeutic outputs. Biology, in short, becomes programmable and conditional rather than merely directed.

For example, T cells engineered with multi-input logic gates can activate only when detecting tumor antigen AND immune checkpoint ligand, avoiding off-target toxicity that has plagued earlier approaches. Or macrophages programmed to detect and respond to immunosuppressive signals, converting from pro-tumor to anti-tumor phenotypes based on local conditions. Synthetic biology enables therapeutic cells that adapt to tumor evolution, adjusting their function as the tumor changes tactics.

We would be fighting adaptation with adaptation—a fair contest at last.

The technical challenges are substantial—genetic circuit design, cell manufacturing at scale, safety controls preventing runaway activation—but the enabling technologies are advancing rapidly in adjacent fields. Cancer therapy may evolve from administering drugs to deploying biological systems that operate autonomously within patients, making therapeutic decisions at spatial and temporal resolutions that no current therapies could match. The implications are simultaneously exciting and mildly disconcerting, as they should be when contemplating autonomous therapeutic systems.

The environmental dimension

Climate and environmental exposome effects on cancer represent an underappreciated frontier, perhaps because it requires uncomfortable conversations about systemic factors beyond individual biology. While tobacco and industrial carcinogens receive sustained attention, broader environmental factors—air pollution, water quality, microplastics, temperature extremes, food systems—remain poorly integrated into cancer science.

We have focused intently on the molecular while neglecting the atmospheric.

Epidemiological evidence links air pollution to lung cancer incidence even in non-smokers, a finding that suggests our ancestors’ concerns about “bad air” were not entirely misplaced, merely imprecisely formulated. Climate change alters exposure patterns in ways both direct and indirect: expanding geographic ranges of carcinogenic pathogens, changing agricultural practices affecting dietary exposures, increasing extreme weather events disrupting cancer care delivery. Long-term exposure modeling requires integrating geospatial, climate, and health data at population scales—a computational and conceptual challenge of considerable magnitude.

This area demands interdisciplinary collaboration between oncologists, environmental scientists, climate researchers, and epidemiologists—groups that currently occupy largely non-overlapping professional and conceptual spaces. The payoff extends beyond research: understanding environmental determinants enables prevention at population levels, potentially reducing cancer incidence more effectively than any therapeutic intervention. Prevention, it bears noting, remains vastly cheaper than cure—though substantially less lucrative, which may partially explain the funding differential.

Materials meet medicine

Advanced biomaterials and nanotechnology offer us precise control over drug delivery, tumor biosensing, and cancer modeling through materials engineered at scales where physics and chemistry behave counterintuitively. Nanoparticles can be designed to circulate inertly until encountering specific tumor microenvironmental conditions—pH, enzymes, hypoxia—triggering drug release. This can theoretically achieve high intratumoral concentrations while minimizing systemic toxicity, solving the perennial problem of getting drugs where they should be and keeping them from where they should not.

Beyond delivery, biomaterials can enable biosensing: implantable devices continuously monitoring tumor biomarkers, providing real-time disease tracking without repeated blood draws. Tissue-engineered tumor models—organoids in defined biomaterial scaffolds—can create experimental systems that recapitulate tumor architecture and microenvironment more faithfully than conventional cell culture, improving drug screening predictive value.

Plastic dishes, whatever their experimental convenience, do not closely resemble human tumors in geometry, mechanics, or microenvironmental complexity.

The materials science community has developed sophisticated capabilities—stimuli-responsive polymers, self-assembling nanostructures, biocompatible hydrogels with tunable properties—that remain substantially underutilized in biomedicine. Bridging this gap requires cross-disciplinary teams fluent in both cancer biology and materials engineering, a combination currently rarer than one might hope given the potential applications.

Quantum leap in computation

Quantum computing for molecular modeling remains speculative but potentially consequential. Classical computers struggle with problems requiring exploration of vast combinatorial spaces—protein folding, drug-target binding prediction, molecular dynamics of complex systems with many interacting components. Quantum computers, exploiting superposition and entanglement, could solve certain problems exponentially faster than classical approaches. Whether they will actually do so in biomedically relevant timeframes remains, appropriately enough, in superposition until measured.

Applications in cancer science include: simulating protein conformational changes to predict drug binding with unprecedented accuracy, screening vast chemical libraries for novel therapeutics, and modeling complex biological networks—signaling pathways, metabolic networks, gene regulatory circuits—at scales impossible with classical computation. The potential is considerable; the timeline is uncertain; and the hype is substantial.

Quantum computing remains early-stage, with current devices limited in qubit number and coherence time, prone to errors, and requiring temperatures approaching absolute zero. But the trajectory is clear, and we should track developments rather than dismissing them as distant speculation. When quantum advantage is achieved for biomedically relevant problems—likely within the next decade or so—early adopters will gain substantial competitive advantage.

Being fashionably late to technological inflection points is rarely advisable.

Movement IV - Andante Moderato: Orchestrating the future

On integration across temporal horizons

The priorities I’ve outlined do not operate independently but form an interconnected ecosystem requiring coordination rather than mere coexistence. Established areas provide the biological foundations and clinical imperatives. Emerging technologies offer new capabilities for addressing longstanding questions that previously resisted solution. Future directions expand what is possible beyond current constraints. The art lies in conducting them simultaneously without cacophony.

As an example, therapeutic resistance, an established priority, can be addressed by tools within reach—spatial multi-omics revealing geographic resistance patterns within tumors, single-cell profiling tracking clonal evolution under treatment pressure, liquid biopsies detecting emerging resistance mutations before clinical progression, and AI integrating multimodal data for predictive modeling. Future capabilities—synthetic biology creating adaptive therapies that respond to resistance in real-time, quantum computing predicting resistance mutations before they occur—will further enhance our capabilities.

The question is not whether tools exist, but whether we can wield them in concert rather than sequentially.

I think the strategic imperative is balance: maintaining investment in established areas while embracing emerging and future capabilities without starving either. This requires institutional flexibility—rare in academic and funding structures optimized for stability rather than adaptation—funding mechanisms that support interdisciplinary teams beyond pilot-scale efforts, training programs producing researchers fluent across traditional boundaries, and patience.

Transformative science operates on decade timescales. Institutions, unfortunately, often operate on quarterly ones, and funding cycles rarely exceed five years.

Finale - On translation and ambition

Cancer research has achieved substantial progress: diseases once uniformly fatal now have effective treatments, and some cancers approach chronic disease management or even cure. Yet the majority of advanced cancers remain incurable, and disparities in outcomes persist across populations in patterns that reflect social structures as much as biological variation.

We have climbed far, but the summit remains distant and occasionally obscured by weather.

The roadmap presented here—reinforcing established priorities, integrating emerging capabilities, and preparing for future technologies—may offer a framework for continued progress. Success requires not choosing between these horizons but orchestrating them into a coherent whole, recognizing that today’s future technologies become tomorrow’s established infrastructure and the day after tomorrow’s legacy systems requiring maintenance or replacement.

The next decade will be defined by integration: of data types, of biological scales, of disciplinary perspectives previously separated by departmental walls and professional incentives. Tumors are complex adaptive systems, and understanding them requires similarly complex, adaptive research ecosystems. The tools increasingly exist. The challenge is increasingly deployment, translating capability into knowledge, and knowledge into improved patient outcomes.

Technology without biology is mere gadgetry; biology without technology is increasingly unimaginable; both without clinical translation are academic exercises.

That translation—from scientific insight to clinical benefit—remains the ultimate measure of success, the metric that survives changes in fashion, funding, and institutional priorities. All technical sophistication, all computational power, all molecular detail must ultimately serve a single goal: reducing the burden of cancer for patients and populations. Keeping that goal central, while embracing the full spectrum of scientific possibility, defines the path forward.

The future of cancer research will be written by those who can master established knowledge while remaining open to disruption, who can work across disciplines while maintaining depth, and who can embrace uncertainty while pursuing rigor. It is, admittedly, a demanding job description requiring contradictory qualities. But then, cancer has never been particularly accommodating to our preferences or convenient for our experimental designs. The very least we can do is return the favor through work that is equally uncompromising in its demands.

Subscribe now

I’ve been thinking about something Mike Krzyzewski (Coach K) said recently.

He told the story of his first night coaching the U.S. Olympic basketball team—Kobe Bryant, LeBron James, the most gifted players in the world gathered in one room. Instead of outlining plays, he gave them a single, unexpected assignment:

“When you go back to your suites tonight, put on your USA uniforms and be sixteen again—for one minute.”

Later that night at the Wynn Las Vegas, Kobe tapped on his shoulder.

“Coach, I did what you asked,” he said.

“What happened?”

“I cried.”

It’s a simple story, but it captures something essential about leadership. People commit when they reconnect with the why—the purpose that first moved them to act—not just the what they’ve been told to do.

That same clarity of purpose is what built the CEO Roundtable on Cancer (CEORT), where I’ve had the privilege of serving as the CEO for about 2 years.

In 1998, President George H. W. Bush looked at the fragmented landscape of cancer research and saw “towers of excellence”—academic centers, biopharma companies, and advocacy groups—operating in silos. “What we have,” he said, “isn’t dialogue. It’s a collision of monologues.” His solution was to bring together the CEOs of these sectors, not to compete but to collaborate.

He appointed Bob Ingram , then CEO of GlaxoWellcome, as founding chairman and paired him with Dr. Martin (Marty) Murphy, a scientist and systems builder who became the Roundtable’s first CEO. Bush’s charge to them was characteristically direct: Be bold and venturesome. Cancer was intolerable; do something about it.

From the beginning, the Roundtable was an experiment in collective leadership.

Bob Ingram invited his friend Coach K to join the founding circle. Bob believed that the coach who had built one of the most cohesive teams in sports could teach CEOs something about aligning strong egos around a shared goal. Coach K accepted—and wore the Roundtable’s lapel pin to every game he coached. He forgot it only once. The next morning, President Bush called Marty Murphy from Houston. “Marty,” he said, “what did you do wrong?”

That attention to detail was anchored in the expectation that if you were part of this mission, you represented it—publicly and consistently.

Years later, when someone proposed bringing President Bush to a Duke game, the idea seemed impossible. His Parkinson’s had advanced; he rarely left home. But when asked, he said, “No, you’re not crazy. I’m giddy as a schoolboy.” With help from Dr. Jim Goodnight, co-founder of SAS and CEORT board member, a jet was arranged from Houston to Durham. When President Bush’s wheelchair entered Cameron Indoor Stadium, the crowd erupted: “U-S-A 41! U-S-A 41!” After the game, Coach K brought him into the locker room and saluted his former commander-in-chief. It was a moment that reflected CEORT’s DNA—commitment meeting commitment.

That same ethos defines our work today.

Chris Viehbacher, now CEO of Biogen and CEORT board member, was one of the original architects of Project Data Sphere, the Roundtable’s data-science and AI subsidiary created to dismantle the same silos President Bush identified decades ago. Chris has a sharp taxonomy: ninety percent of people complain; ten percent do something; one percent mobilize others. Mobilizing others is the hard part—and the essential one.

At Project Data Sphere, mobilization takes the form of shared data environments where competitors collaborate safely on pre-competitive challenges; of AI systems that quantify tumor response more precisely than legacy methods; of frameworks that connect innovation to access. It’s how we turn purpose into structure.

Coach K once told me that when he coached the U.S. team, he didn’t ask players to leave their egos at the door. He asked them to bring them inside—then align them under a single banner. “You’re in it to win,” he said. “And it doesn’t make sense that you win alone.”

That philosophy inspired something new: Precision Signals, CEORT’s signature video podcast series .

Precision Signals is a platform for substantive dialogue across disciplines—a continuation of the Roundtable’s founding premise that progress happens when silos dissolve. It’s where I have the privilege of speaking with the people and ideas transforming biomedical research and healthcare: scientists, clinicians, technologists, investors, government officials, and policymakers whose work embodies the same spirit of bold collaboration .

Each conversation explores how insight becomes evidence, how evidence becomes policy, and how innovation becomes access. In a recent episode, I spoke with Dr. James Gulley, Co-Director of the Center for Immuno-Oncology at the National Cancer Institute, about the evolving frontier of cancer immunotherapy and the role of intelligence—both human and artificial—in shaping its future. It’s precisely the kind of exchange we aim to surface: where bold ideas meet strategic action.

“You can’t win alone” is an organizing principle . It’s what connects a locker-room lesson in Las Vegas to a national experiment in collective leadership, and now to a global dialogue about the future of science and healthcare.

To follow these conversations and the people driving them, subscribe to Precision Signals—and join us as we keep tracing the signal beneath the noise.

Subscribe now

It was a great pleasure to sit down with renowned investor and social entrepreneur Daniel Arbess for a deep conversation about the future of healthcare and biomedical research. Dan’s career trajectory—from foreign policy analyst to the youngest partner at White & Case, to founding Xerion (a $3.5 billion hedge fund that delivered 25%+ annualized returns through the financial crisis), to his current work in patient advocacy—offers a unique lens on system-level transformation.

Our conversation centered on a fundamental thesis: healthcare doesn’t just need more science; it needs a moonshot of institutional innovation. Dan argues that our current system was architected 100 years ago to treat visible symptoms with available therapeutics. We now have the scientific capacity to detect disease at its molecular origins and treat patients as true N-of-1s, but our institutions haven’t evolved to match.

Several key points emerged from our conversation:

On regulatory reform: Dan argues that if we accept every patient is biologically unique, proving interventions work uniformly across populations becomes less relevant than allowing physicians to apply treatments based on mechanistic understanding, pointing to the need for FDA flexibility in making approval decisions.

On data infrastructure: The path forward begins with standardized electronic health records that capture longitudinal patient data from the start—not after disease manifests. This creates the foundation for understanding individual phenotypes and their evolution over time.

On economic alignment: The current fee-for-service model and diagnostic billing codes perpetuate symptom-focused care. Dan sees the current level of federal budget spent on healthcare as recoverable through early intervention and prevention, rather than managing expensive end-stage disease.

On the opportunity: With an administration open to disrupting calcified systems, there’s a unique window to redesign how we generate evidence, reimburse care, and ultimately practice medicine.

Dan’s work in neuroimmunology—looking across Alzheimer’s, Parkinson’s, ALS, and brain tumors to identify common dysregulatory mechanisms—exemplifies this systems-thinking approach. His first academic presentation at the Society for Neuro-Oncology (which we collaborated on) attempted to extract learnings from failed trials to inform better trial design. The limited feedback he received underscores the challenge: the field continues designing studies without fully integrating what we’ve learned about heterogeneity and mechanism.

The conversation reinforced something I’ve been thinking about: the gap between our technical capabilities and our institutional capacity to deploy them may be the defining constraint in modern medicine. We have the tools. We lack the operating system.

Subscribe now

The United States continues to spend more on health care than any other country, yet the relationship between expenditure and outcome remains stubbornly misaligned. National health expenditures (NHE) reached $4.9 trillion in 2023, or 17.6% of GDP, marking the fastest growth since the acute pandemic years. Forecasts anticipate further acceleration—7.1% in 2025—placing U.S. health care on a trajectory well above pre-COVID baselines. Despite this expansion, U.S. life expectancy in 2023 stood at 78.4 years, a modest recovery from pandemic lows but still below the 2019 level of 78.8 years. The paradox is rather simple: expenditures climb, but improvements in health outcomes are marginal and in some cases deteriorating.

I believe the persistence of this imbalance is not due to a lack of clinical expertise or capital—America’s physicians and clinical workforce re among the best in the world and the sector is well-resourced—but rather to how resources are structurally allocated. While health systems and payers often cite competing investment priorities, the real issue is in the frameworks and incentives that shape these decisions and determine whom they benefit. Comparative analyses consistently show that the U.S. spends disproportionately on administration. Studies estimate administrative functions—spanning billing, coding, eligibility verification, payer overhead, and compliance—as consuming 15–25% of total expenditures, or approximately $1 trillion annually at current levels. By contrast, peer nations direct a significantly smaller fraction of resources to non-clinical functions while achieving longer life expectancy and lower preventable mortality. This suggests not a scarcity of resources but a clear misdirection of investment.

The infrastructure crisis

Cancer care and research offer clear examples of the consequences of resource misallocation. Despite ongoing scientific advances, participation in cancer clinical trials remains low: recent estimates suggest that at most ~7% of adult cancer patients in the U.S. enroll in treatment trials—far below the level needed to translate innovation into meaningful patient impact through clinical research and evidence generation. The barriers are largely infrastructural: trial sites are geographically concentrated, and interoperable systems for patient screening and consent are lacking. Although investment in clinical research technologies are often discussed in recent years, the foundational elements—such as digital pre-screening integrated into community EHRs, standardized eConsent, and reimbursement for trial navigation—are still sporadically implemented, leaving real progress elusive.

This challenge extends far beyond clinical trials and care delivery, with significant implications for the broader adoption of AI-driven technologies in biomedicine. The promise of artificial intelligence—from advanced diagnostics to workflow automation—depends on robust infrastructure capable of supporting seamless data integration, reliable storage, and interoperable clinical workflows. Digital pathology is a telling example: although the FDA has cleared whole-slide imaging for primary diagnosis and a growing ecosystem of AI-assisted diagnostics, routine adoption remains limited. The barriers are not technological capability—digitization is hardly cutting-edge science—but rather financial and organizational priorities. High storage costs, inconsistent reimbursement, and capital budgets focused on immediate throughput rather than long-term infrastructure continue to impede widespread implementation and, as a result, the full realization of AI’s transformative potential in clinical care.

The evolution of cancer therapeutics highlights just how far our infrastructure has lagged behind clinical innovation. Modern oncology now extends well beyond the infusion chair: subcutaneous biologics allow for home-based treatment, oral targeted therapies shift care from the clinic to the patient’s home, and antibody-drug conjugates offer unprecedented precision but require complex, multidisciplinary coordination. Their side effects can span multiple organ systems—ocular toxicities require immediate ophthalmologic assessment, cardiac effects call for cardiology intervention, and dermatologic reactions demand prompt attention from specialists. Yet, our current infrastructure is built on outdated assumptions: that all cancer care happens within specialized centers, that oncologists manage every aspect, and that referrals can wait weeks. In reality, without structured pathways for rapid specialist access and reimbursement for multidisciplinary monitoring, patients are left with fragmented, ad hoc care that undermines both efficiency and safety. Our science has leaped forward; the systems to deliver it have not.

The scale of misallocation

The scale of potential resource reallocation is rather striking. Redirecting even one percentage point of NHE—roughly $50 billion in 2024 terms—from administrative overhead to infrastructure could fund nationwide digitization of pathology, build interoperable trial access networks, and establish compensated multispecialty toxicity management frameworks. The measurable outputs would include higher trial accrual velocity, reduced diagnostic latency, and fewer avoidable emergency department visits. These are not speculative benefits but quantifiable improvements in efficiency and outcomes.

Capital is by no means the binding constraint. Private equity and venture capital firms currently hold "mountainous and aging" levels of dry powder, with deployment slowed by narrower deal theses and herd behavior around familiar asset classes. The reluctance is structural: shared infrastructure that produces sector-wide positive externalities does not fit neatly into conventional fund-return models. Yet analogous sectors at inflection points—information technology during the internet protocol standardization of the 1990s, or aviation during the development of interoperable safety protocols—achieved successful infrastructure build-out through coalition funding and public–private incentives, delivering substantial returns for those playing the long game.

The provocation

With national health expenditures nearing $5 trillion, we should expect more than administrative complexity and modest, incremental improvements in outcomes. Redirecting spending toward tangible, verifiable infrastructure may not yield immediate cost reductions—real progress is a long game—but it would transform financial outlay into measurable gains in health. History shows that those willing to invest with patience and vision can realize substantial returns, both in value and in impact.

We have built an administrative state in the biomedical enterprise that consumes resources while producing paperwork. Every administrator hired to navigate complexity is a clinician not hired to reduce suffering. The American health care system now stands at a true inflection point.

We can continue to feed the administrative apparatus—layering on complexity and cost while outcomes stagnate or deteriorate. Or we can choose to redirect that trillion-dollar stream toward infrastructure that actually improves health: diagnostic networks that catch disease early, trial platforms that match patients to breakthroughs, and coordination systems that ensure comprehensive care. The alternative is more of the same—expecting different results from the same misdirected investments.

The choice is not about spending more or less, but about spending wisely and intentionally. The question is urgent and unavoidable: What are we spending our money on—and what do we want that investment to achieve?

Subscribe now

Immunotherapy has transformed the landscape of oncology, but many of its scientific and structural challenges remain unresolved.

In this episode of Precision Signals, I sat down with Dr. James Gulley, Chief of the Medical Oncology Service and Co-Director of the Center for Immuno-Oncology at the US National Cancer Institute. Dr. Gulley has been at the forefront of immuno-oncology for decades—helping move immune-based therapies from early skepticism to clinical reality while also shaping the frameworks that govern their evaluation and approval. I’ve had the privilege of working with Dr. Gulley across multiple domains, from the early development of immunotherapies to efforts integrating AI into clinical research.

At each inflection point in oncology, Dr. Gulley has brought a systems-level perspective that links mechanistic insight with operational execution, and scientific rigor with institutional change. Whether developing adaptive clinical trials or reimagining evidence generation and real-time data capture with machine learning, his work reflects a sustained commitment to redefining what translational and clinical research can achieve—and what it should require of us.

Our conversation traces the field’s evolution from first-in-human trials to today’s adaptive, biomarker-driven studies. We examine the limits of predictive biomarkers, the complexity of resistance mechanisms, and the inadequacies of current trial designs. We also discuss the need for new endpoints and evidence hierarchies that reflect patient heterogeneity more faithfully.

We also explore how AI-enabled protocol generation, wearable-derived physiologic monitoring, digital pathology, and EHR-integrated data flows are beginning to reconfigure translational infrastructure. Dr. Gulley shares insights from recent work using continuous monitoring to detect CAR-T toxicity before it manifests clinically—an example of how next-generation research tools are converging with real-world care.

Share

Subscribe now