Part I
A bet on simulation
“Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.”
McCarthy, Minsky, Rochester & Shannon · Dartmouth proposal · August 31, 1955
The 1955 Dartmouth proposal stated a conjecture: anything described precisely enough can be simulated by a machine. The same conjecture now underlies computational biology: if the protein, the cell or the patient can be described precisely enough, drug discovery becomes an engineering problem.
The history of AI includes two “winters,” periods of reduced funding and interest, each following a phase in which expectations exceeded what the methods could deliver.
Part II
The molecule that could not
RG7112 was designed to free the tumor suppressor p53 by blocking MDM2, the protein that holds p53 in check. Many tumors keep p53 silent this way, so releasing it is a direct strategy: restore the cell’s own brake.
The molecule was developed with the best tools available at the time and cleared each preclinical test. Step through its development to see where it failed.
-
Binding assays identified the nutlins, the first selective small-molecule antagonists of MDM2.
Vassilev et al., Science 2004
-
X-ray crystallography showed how the molecules sit in MDM2’s binding pocket (PDB 1RV1; RG7112 bound, 4IPF). In a binding assay, RG7112 reached an IC50 of 0.018 µM.
Vu et al., ACS Med Chem Lett 2013
-
Structure-guided chemistry tuned potency and metabolism.
Vu et al., ACS Med Chem Lett 2013
-
Mouse pharmacokinetic modeling showed about four times the exposure at half the dose, with a half-life of 8.8 hours.
Vu et al., ACS Med Chem Lett 2013
-
In SJSA-1 osteosarcoma xenografts, RG7112 inhibited tumor growth by 74% at 50 mg/kg and caused regression at 100 mg/kg.
Vu et al., ACS Med Chem Lett 2013
-
Toxicology studies supported a favorable safety profile. Every preclinical gate was cleared.
Vu et al., ACS Med Chem Lett 2013
-
In patients with MDM2-amplified liposarcoma, the drug engaged its target and activated p53 as designed. Efficacy was minimal and hematologic toxicity was dose-limiting.
Benefitp53-pathway activation confirmed in tumor biopsies. Only one confirmed partial response.
HarmSerious adverse events, including neutropenia and thrombocytopenia.
Ray-Coquard et al., Lancet Oncol 2012
It hit its target and missed the patient.
Part III
New tools of the trade
“For the very first time in human history, biology has the opportunity to be engineering, not science.”
Jensen Huang, 2023
A decade later, four families of models cover most of the steps RG7112’s chemists performed by hand.
Predicting how molecules fit together
AlphaFold 3
Predicts the joint structures of complexes: proteins, nucleic acids, small molecules and more.
Abramson et al., Nature 2024
RFdiffusion
A diffusion model for de novo design of protein structure and function, such as new binders.
Watson et al., Nature 2023
Reading the genome’s regulatory code
AlphaGenome
Takes 1 Mb of DNA in and predicts thousands of functional tracks at up to single-base resolution. Matches or beats the best external models in 25 of 26 variant-effect evaluations.
Avsec et al., Nature 2026
Evo 2
40 billion parameters, a 1 Mb context, trained on more than 9 trillion nucleotides. Predicts and designs across DNA, RNA and proteins.
Brixi et al., Nature 2026
Designing proteins not found in nature
ESM3
A multimodal generative language model that reasons over protein sequence, structure and function. It produced a bright new fluorescent protein only 58% identical to known ones: a distance equivalent to about 500 million years of evolution.
Hayes et al., Science 2025
Language models join the lab
TxGemma
Language models (2B, 9B and 27B) that predict therapeutic properties with natural-language reasoning. They match or beat the best generalist model on 64 of 66 tasks. Wet-lab validation is still a future step.
Google DeepMind, 2025
AI co-scientist
A multi-agent system built on Gemini 2.0 that generates novel hypotheses and research proposals.
Google, 2025
Part IV
Would today’s tools catch it?
“It takes ten years and billions of dollars to design just one drug. We can maybe reduce that down from years to maybe months or maybe even weeks.”
Demis Hassabis, 2025
The main claim is speed. Early clinical results are favorable through Phase I; the advantage is not evident in Phase II, where efficacy is first tested.
Analysis of AI-native biotech pipelines · Jayatunga et al., Drug Discov Today 2024
Phase I mostly tests whether a molecule is safe enough to dose. Phase II is where it first has to help patients, and that is where the advantage disappears. To see why, run RG7112 through today’s toolkit.
Within the models’ view
- Binding pose in MDM2 from AlphaFold-class structure prediction
- Potency and drug-like properties from property models
Outside it
- On-target p53 activation in bone-marrow progenitors, which led to neutropenia in 6 patients and thrombocytopenia in 3
- The therapeutic window in humans
The toxicity was on-target: RG7112 activated p53 in bone-marrow progenitors, a human tissue context the design models did not represent. Ray-Coquard et al., Lancet Oncol 2012
Part V
Why Eroom’s law won’t break
Eroom’s law, Moore’s law spelled backward: since 1950, the number of new drugs approved per billion US dollars of R&D has halved roughly every nine years: about an 80-fold decline, adjusted for inflation, despite large advances in science, technology and management. Since about 2010 the decline has leveled off, with a modest recovery that Ringel and colleagues attribute largely to better information for choosing targets and a shift toward rare and genetically defined diseases, not to AI. AI-designed molecules have not yet changed the trend.
Schematic of the published trend, not individual data points: an 80-fold decline from 1950 to 2010, inflation-adjusted, halving roughly every nine years. The trend has flattened since 2010. Scannell et al., Nat Rev Drug Discov 2012 · Ringel et al., Nat Rev Drug Discov 2020
What we know often surprises us
Single driver mutations are the exception: most patients with metastatic non-small cell lung cancer and other advanced cancers may not have tumors with a single driver. And even a true driver depends on context.
Feedback signaling through EGFR in colorectal tumors explains part of the gap. Chapman et al., NEJM 2011 · Kopetz et al., J Clin Oncol 2015 · Khozin, Kim & Pazdur, Nat Rev Drug Discov 2017
Our gold standards can be blunt tools
Animal models are the standard gate between the bench and the clinic. In oncology, their predictive value for human toxicity is close to chance.
Animal-to-human toxicity in oncology · Atkins et al., Br J Cancer 2020 · FDA, Roadmap to Reducing Animal Testing, 2025
Part VI
What’s working today
Engineering at full speed
Where the biology is already understood, AI-assisted design is paying off. The clearest signal is in licensing: China-origin assets now account for a large share of big-ticket deals, built by applying fast, AI-assisted design at full scale to validated targets.
J.P. Morgan, Q2 2026 Biopharma Licensing and Venture Report
These molecules are mostly not first in class. They are better versions of known answers, and no new biological understanding is required to make them.
The quiet wins are in the back office
Ambient scribes
Capturing clinical documentation.
Documents
Document management and drafting of regulatory submissions.
Workflow
Automation across business operations.
These tools improve efficiency; they do not add biological insight.
Part VII
Why biology isn’t moving
Four obstacles keep AI’s progress in engineering from turning into progress in biology.
1Fragmentation
Of data. Trials, health records, imaging, omics and outcomes sit in separate silos. A model learns only the slice it can see, and no single slice contains the patient.
Of work. AI itself was the by-product of one conversation across disciplines. In the 1940s and 1950s, neuroscience (McCulloch and Pitts’s neuron as a logic unit, 1943), mathematics (Turing’s work on logic and computability, 1936 and 1950) and computer science (stored programs and the Dartmouth workshop, 1956) were in the same room. Today these fields are highly specialized. Medicine and AI both require multidisciplinary work and multidisciplinary data.
Suppose someone invented a biosensor and wanted to use it to make life-and-death diagnoses.
The reasonable answer today: that’s premature. Convene a workshop. Talk to the FDA.
It is the electrocardiogram.
- 1887Augustus Waller records the heart’s electrical potential from the skin.
- 1903Willem Einthoven’s string galvanometer is ready for use.
- 1905On March 22, Einthoven, a physiologist, and engineer Johannes Bosscha transmit an ECG by telephone line, about 1.5 km.
- 1924Einthoven receives the Nobel Prize.
Physiology and engineering, working together, moved it quickly into practice. The ECG, the X-ray (a physicist’s discovery turned clinical tool) and MRI (physics, chemistry and medicine combined) all came from medicine and engineering sharing a bench. Each would be hard to develop today, across disciplinary and regulatory barriers.
Hjelm & Julius, J Telemed Telecare 2005 · National Academy of Sciences
2Arbitrary gold standards
Human visual assessment is a variable instrument. Much of the ground truth used to train AI comes from human reading of images. In an FDA meta-analysis of 13,677 patients in registration trials, the investigator and an independent radiologist reading the same scans disagreed on tumor response about 30% of the time, and up to 45% in ovarian and pancreatic cancer. Assessments by eye vary within and between observers, in radiology and histopathology alike.
Explore the dataThe full reader-agreement matrix, cell by cell →
Theme 03What this disagreement does to a trial: The Broken Instrument →
The labels themselves are historical accidents. Consider what “lung cancer” means.
“Lung cancer”: an early surgical observation of where a tumor sits.
“Small cell” vs “non-small cell”: what a pathologist sees through a microscope. “Non-small cell” defines a cancer by what it is not.
Distinct molecular drivers, grouped together as one disease.
Train AI on these labels and, at best, it learns to reproduce them faster: with no gain in biological insight.
Khozin et al., FDA meta-analysis, 2017 · Khozin, Clin Transl Sci 2024
Theme 02How cancer got these names, and what replaces them: From Organs to Algorithms →
3Medicine is not SaaS
Ship it
A regulatory clearance feels like a launch. Technology companies and their investors shy away from clinical trials.
Prove it
Medical products need clinical validation. A 510(k) clearance alone does not win clinicians or payers.
The two models carry different evidentiary requirements. Pear Therapeutics, a pioneer of FDA-authorized prescription digital therapeutics, filed for bankruptcy in 2023. Regulatory acceptance of simulation-based evidence has meanwhile advanced faster than industry adoption.
- Oct 2024
FDA Science Board report on new methods
- Apr 2025
Plan to phase out animal testing; AI toxicity data welcomed in IND applications
- Apr 2025
Roadmap: organ-on-chip, computational models, in vitro assays
- Mar 2026
Draft guidance: a validation framework
- Next
No validated simulation submission yet
Regulators now accept simulation-based evidence in defined settings; industry uptake has lagged.
4Limited knowledge of human biology
1–5%?
An illustrative estimate of the share of human biology that has been mapped: a small fraction. It counts only gaps that are already recognized; actual coverage may be lower.
The human genome encodes about 20,000 proteins, many still uncharacterized, and research keeps returning to the ones already well studied. By one systematic ranking of how much is known about each protein family, nearly a quarter of human families (23%) still sit at the very bottom. A simulator cannot simulate what no one has measured.
Rocha et al., PLOS Biol 2023
Part VIII
The way forward
Labels should come from mechanism, not from human sensory input
From organs to algorithms: a bottom-up classification of disease that reduces reliance on subjective human interpretation.
| Dimension | Current classification | AI-driven classification |
|---|---|---|
| 01Primary basis | Anatomical location and histology | Molecular, imaging and histopathological data, integrated |
| 02Precision | Limited by inter- and intra-observer variability | Potentially more objective and reproducible |
| 03Novel features | Limited ability to find new subtypes | Can identify de novo features and patterns |
| 04Biological relevance | May group heterogeneous cancers together | Aims to reflect underlying disease mechanisms |
| 05Treatment matching | Often suboptimal | Potential for more precise selection |
Adapted from Table 1, Khozin, Clin Transl Sci 2024
Theme 02All ten rows, and the case behind them →
From big data to smart data: study the whole patient
Moving from a reductionist center, a single structured table, to a holistic edge that spans the volume, variety, velocity and veracity of data: records, trials, sensors, imaging and omics, joined around one patient.
Explore the figureThe four V’s capacity map, interactive →
Three moves to consider
- 01
Fund mechanism-based labels
Replace labels from the human eye with molecular and multimodal ones.
- 02
Build whole-patient data
Link trials, records, imaging and omics.
- 03
Validate where decisions are made
Measure AI against human outcomes, not benchmarks.
Better molecules are not better medicines.
AI’s biggest opportunity is not designing the molecule. It is discovering the human biology that decides whether it works. RG7112 engaged its target as designed. It failed on human biology the models did not capture.
“We can only see a short distance ahead, but we can see plenty there that needs to be done.”
Alan Turing, 1950
Connections
Where this theme meets the rest of the work.
- Theme 02From Organs to Algorithms: how disease labels arose, and a classification built from mechanism up
- Theme 03The Broken Instrument: what reader disagreement does to trials and the drug pipeline
- Theme 04There Is No Average Patient: trial populations versus real ones, with a reweighting tool
- Theme 05The Patient-Centered Estimand: from hazard ratios to patients helped and harmed
- ResearchInter-reader response classification: the 13,677-patient matrix behind “arbitrary gold standards”
- INFORMEDFrom big data to smart data: the four V’s and the systems view of drug development
- ConvergenceDisease classified by mechanism instead of anatomy, inside a learning system