Glossary

The convergence of evidence: when independent lines of inquiry, often using entirely different methods and resting on different assumptions, arrive at the same conclusion.

A dynamic, continuously updated virtual replica of a biological entity (molecule, cell, organ, patient, hospital, or population) linked to its physical counterpart by a flow of real data, and used to simulate how that entity will behave under different conditions before acting in the real world.

Three features separate a true twin from an ordinary predictive model. It is individualized, built from one patient's multimodal data rather than a population average. It is bidirectionally connected, so that new measurements refresh the twin and the twin's predictions feed back into decisions. And it is generative and counterfactual, existing to answer what-if questions through virtual experiments that would be impossible, unethical, or too slow to run on the person.

Distributed intelligence. Knowledge and decision-making spread across physicians, patients, AI systems, robotics, information repositories, and an expanding cast of healthcare professionals, no longer centralized in one mind.

Horizontal care. Care organized through flattened, networked, role-based teams whose membership shifts by function, in place of vertical hierarchies with fixed authority.

Expertise and decision-making distribute; authority migrates from credentialed knowledge toward interpretation and judgment.

Physician identity ceases to be a fixed role assigned at the outset of training and becomes a dynamic set of coexisting identities and functions held across a career, assembled, set down, and reassembled as the work and the person require.

Care organized around the whole person rather than the presenting complaint, bringing conventional and complementary approaches together, and treating biological, behavioral, social, and environmental domains as one interacting system.

A health system in which science, informatics, incentives, and culture are aligned for continuous improvement and innovation, where best practice is seamlessly embedded in care delivery, patients and clinicians are full participants, and new knowledge is captured as an integral by-product of the care experience.

The defining move is the closing of the loop. Care generates data, data generates evidence, evidence changes care, and the cycle repeats without waiting for a separate research enterprise to intervene.

Virtue. In philosophy (Greek aretē, Latin virtus), a stable excellence of character, a settled disposition, or hexis, as distinct from a fleeting feeling or mere rule-following, that enables a person to perform their proper function well and contributes to flourishing (eudaimonia). In Aristotle's account, moral virtues regulate desire, emotion, and action and are acquired by habituation; each is a mean relative to us, lying between excess and deficiency. Virtue makes both the agent good and the agent's characteristic activity excellent, performed for the right reasons, with the right feelings, at the right times.

Medical virtue. That excellence specified to the ends of medicine: the settled dispositions required for the physician to achieve the proper ends of the practice, namely the patient's medical good, their experienced good, their good as a person, and their spiritual good. Not optional traits of temperament but constitutive requirements of competent practice, derived from the nature of medicine as a moral community rather than imported from outside it. Canonically: fidelity to trust, compassion, prudence, justice, courage, temperance and self-effacement, and integrity.

Virtue is not niceness and not rule-observance; it is the excellence that makes a practitioner good at being what they are.

Ultra-personalized medicine coupled to high-dimensional diagnostics: the ambition of treating each patient as their own cohort instead of as a member of one.

The move abolishes the subgroup; it does not narrow it further. Population medicine asks what works on average. Stratified medicine asks what works in this class of patient. N-of-1 medicine asks what works in this person, accepting that the reference class may contain a single member. It requires dense, continuous, multimodal measurement, and a means of reasoning from mechanism when no comparable case exists.

One Health. A unifying, multisectoral approach recognizing that the health of humans, animals both domestic and wild, plants, and the wider ecosystem is interdependent, and aiming to balance and optimize all three. Its practical focus has been zoonoses, emerging infection, food security, and antimicrobial resistance. The idea is old: Virchow, who coined zoonosis, held that between animal and human medicine there is no dividing line. Its proximate source is the veterinary epidemiologist Calvin Schwabe, who introduced One Medicine in 1964; the current term was popularized through the twelve Manhattan Principles agreed at Rockefeller University in 2004, and institutionalized through the One Health Initiative later that decade.

Planetary Health. The health of human civilization and the state of the natural systems on which it depends, centered on how human disruption of Earth systems threatens health, with heavy emphasis on equity and on future generations. Its origin is the Rockefeller Foundation-Lancet Commission report Safeguarding Human Health in the Anthropocene Epoch (2015), preceded by the 2014 manifesto From Public to Planetary Health, and consolidated by the Planetary Health Alliance and The Lancet Planetary Health.

One Health looks outward, across species at the human-animal-environment interface. Planetary Health looks forward, at Earth's life-support system and its intergenerational stakes: it is distinctively concerned with time. They are complementary siblings, not synonyms.

Durable configurations of capability, disposition, and role that a physician assumes, held plurally and in varying proportion across a career, and overlaid on specialty.

Machine intelligence that decisively exceeds the best human performance across essentially all cognitively demanding domains, rather than in a narrow task.

The term carries several readings. Speed superintelligence runs human-equivalent reasoning far faster; collective superintelligence aggregates many systems into a whole that outperforms any one of them; quality superintelligence reasons better in kind. A looser usage treats it as anything beyond artificial general intelligence.

Relation to cognition. The term is defined by cognitive performance, which is where the disagreement sits. On a functionalist reading, cognition simply is what a system does when it represents, infers, plans, and revises, so a system that does these better than any human is more cognitive, and superintelligence is attainable in principle. On the rival reading, cognition also requires understanding and first-person experience, which performance alone cannot establish, so a system might surpass humans without attaining cognition at all. A third view holds the question undecidable from the outside.

A gap rooted in theorems, physical principles, or deep structural features of knowledge and reality itself, not a temporary engineering shortfall that better algorithms, materials, or funding will close.

The distance between what biomedical science has made translatable and what has actually been translated into practice, measured not in discovery, but in delivery.

The deliberate multi-modality integration of diagnostic and therapeutic functions into the same platform (molecule, device, or system) for a specified deliverable.

The word is often used loosely to mean any pairing of a test with a treatment. The stricter sense used here is more demanding. Diagnosis and therapy are not merely sequential but co-located, so that the act of sensing and the act of intervening occur within one engineered system, and each informs the other in something close to real time.