Novartis MAIO

Who delivered the guest lecture on P4+AI for better health in April 2026?

Michael Rebhan, Ph.D. from Novartis.

According to Thomas Kuhn's "The Structure of Scientific Revolutions," what happens during a scientific paradigm shift?

An old paradigm experiences stability, anomalies accumulate, heated debates erupt between defenders and enthusiasts, a new generation adopts a new stable framework, and the old paradigm fades out.

What are Katri Valkokari's 3 ecosystem types referenced in the pipeline view of paradigm shifts?

Knowledge Ecosystems (generating new knowledge/tech), Innovation Ecosystems (integrating exploration and exploitation), and Business Ecosystems (creating customer value).

Trace the historical paradigm shifts in drug discovery and medicine mentioned in the lecture.

Herbs & spirits (Holistic health) → Galenic medicine/Blood letting → Alchemy → Microscope (Western science/tech) → DNA & High-tech (Western/Traditional medicine).

What are the 4 AI archetypes and subcultures mentioned for organizational deployment?

LLMs (ChatGPT, Llama2, DeepSeek), Imaging AI (histology/medicine), Beyond LLMs & Imaging AI (operational deployment), and "Classic" data science and statistics.

What subcultures evolved in the modern era of AI history after the "AI winters"?

Machine Learning, Deep Learning ("black box"), Learned Representations, RAGs, Neuro-symbolic AI, Computer Vision, NLP, Robotics, and Multi-agent systems.

What historical diagnostic system represents the "Age of Expert Systems" (1973-78) in AI history?

MYCIN.

What core challenges exist when moving a new AI model from development to operational readiness?

Transitioning it into an organization of humans—ensuring it delivers value, fits operationally, handles human/management dynamics, and mitigates unforeseen side effects.

What does a Responsible AI framework with Human-in-the-Loop (HITL) entail?

Validation of training data by humans, human-led AI model correction, and validation of system outputs by humans to establish operational accountability.

What mechanism did OpenAI use to align ChatGPT with human preferences?

RLHF (Reinforcement Learning from Human Feedback).

What do the four pillars in "P4 Medicine" stand for?

Predict (upcoming health issues), Prevent (bad outcomes), Personalize (healthcare), and Participate (neglected voices).

Contrast the P4 perspective with the Longevity ecosystem focus.

P4 emphasizes health maintenance and proactive preventive care; Longevity focuses on healthspan, biological clocks vs. chronological age, luxury health checks, and the science of aging hallmarks.

Using the individualist perspective, what layers of Maslow's hierarchy must align with a patient's health engagement?

Physiological, Safety, Love/belonging, Esteem, and Self-actualization.

What metric approximates kidney health, and what are its baseline numbers?

eGFR (estimated Glomerular Filtration Rate): 90 indicates healthy function, while 15 represents severe kidney status or failure.

Why is a "fast decline" in kidney health problematic in traditional healthcare systems?

By the time a fast decline becomes obvious through strong clinical symptoms, it is often too late to safely intervene, leading to dialysis or transplantation.

Apply the P4 principles directly to the Chronic Kidney Disease (CKD) exemplar.

Predict the eGFR trajectory, Prevent rapid decline early, Personalize the prevention approach, and use Participation to make it work in real life.

What is the primary economic challenge of deploying a P4 medicine approach?

The value "flow-back" time frame, as preventive investments require demonstrating decreased uncertainty, increased trust, and clinical evidence before achieving ROI.

Contrast the classic ways of organizing (circa 1200 AD) with modern AI/P4 ecosystems.

Classic systems rely on territorial power, rigid hierarchies, centralized orchestration, and rigid frames (e.g., bloodletting); P4/AI requires open, less territorial, and collaborative cross-organizational mindsets.

Give an exemplar of decentralized orchestration using digital blockchain assets in health sciences.

VitaDAO, which coordinates DAO members, VITA tokens, IP NFTs, and data assets to fund labs and license intellectual property to third parties.

What real-world health partnership in Switzerland serves as a P4+AI operational model?

The integrated care partnership between SMN (Swiss Medical Network), Visana health insurance (the Viva insurance model), and the canton of Berne.

What operational framework brings the science of P4 into the Swiss Medical Network's operational care?

The 8P model, which introduces new roles to fill organizational competency gaps related to P4+AI.

What are the 3 P4 Ecosystem Archetypes on a growth path today?

The Integrated Care Archetype (e.g., Réseau de l'Arc/SMN), the Omics Archetype (e.g., Nightingale Health), and the Connected Health Archetype (e.g., HiNouNou).

Describe the design and economics of the Integrated Care Archetype.

It combines classic integrated care design with P4 extensions, relying on capitation models and risk-sharing between local governments and health insurance.

Describe the design and deployment of the Omics Archetype.

It deploys advanced omics technology in the real world by partnering with organizations like employers who care about workforce health, yielding fast value.

Describe the Connected Health Archetype.

It utilizes digital technology for health state home monitoring by partnering with elderly care, regional governments, or telecom operators, requiring heavy local customization.

What is the primary focus of AI traction in modern P4 ecosystems?

Predicting health declines, organizing smarter resource-limited prevention, driving organizational/ecosystem-level learning, and managing risk without massive traditional RCT evidence.

What core vision from Tim Berners-Lee (1999) underpins the shift beyond basic knowledge graphs?

The Semantic Web vision (using RDF structures), where semantics are fully understood by both humans and AI agents to guide unguided patients through complex disorders.

What is the definition of "<>" in learning health systems?

The return loop where data collection decreases clinical or operational uncertainty, directly translating into better healthcare decisions.

Complete the evolution phrase: "MDs who work well with AI will replace those who don't; what's next?"

MDs who understand P4+AI at the ecosystem level, recognizing there is a massive health world completely outside the hospital walls.