What is augmented reasoning?

Augmented reasoning is a deliberate, structured way of using generative AI — large language models such as Claude, GPT, Gemini, and others — not as an answer machine, but as a reasoning partner for an expert or a panel.

• The human expert drives an iterative, interactive process: formulates the question, chooses the analytical framework, evaluates the output, and iterates.
• The AI contributes breadth of knowledge, systematic exploration of alternatives, and tireless patience for "what if" scenarios.
• The result is a form of thinking that neither the human nor the machine could achieve alone.

This approach is distinct from the more familiar uses of AI in healthcare (diagnostic imaging, drug discovery, predictive analytics).

Augmented reasoning targets the cognitive work of healthcare professionals: designing care pathways, writing policy documents, analysing complex multi-stakeholder scenarios, comparing standards and frameworks, planning integrated care initiatives, and making sense of large bodies of evidence.

Why this community?

Healthcare is facing challenges that are not primarily computational but cognitive: how to integrate fragmented services, how to design person-centred pathways across multiple organisations, how to reconcile clinical standards with local realities, how to plan long-term care for ageing populations with multiple chronic conditions.

These are reasoning-intensive tasks that could benefit enormously from a structured human–AI collaboration.

This community was created to:

• Share concrete, reproducible examples of augmented reasoning applied to real healthcare problems — not toy demos, but actual working sessions with real prompts, real outputs, and real reflections on what worked and what did not.
• Propose a methodology for using AI as a reasoning partner, including prompt design patterns, analytical frameworks (Soft Systems Methodology, lateral thinking, multi-perspective analysis), and quality criteria for evaluating AI-assisted reasoning.
• Invite open discussion among healthcare professionals, health informaticians, policy makers, and researchers who are experimenting with similar approaches or who want to start.
• Build a shared evidence base that can inform both practice and policy on the responsible, effective use of AI in healthcare planning and management.

What you will find here

The community hosts a growing collection of deposits including:

• Methodological notes explaining the augmented reasoning approach, its theoretical foundations, and its relationship with established methodologies (Soft Systems Methodology, clinical reasoning, lateral thinking).
• Worked examples (annotated prompt–response sequences) showing how augmented reasoning has been applied to specific healthcare problems: designing integrated care pathways, analysing the European Health Data Space (EHDS), comparing clinical information standards (openEHR, FHIR, HL7 CDA), planning territorial care networks, and more.
• Analytical papers exploring specific topics in depth through augmented reasoning, with full transparency on the AI tools used, the prompts given, and the human editorial process.
• Discussion papers and commentaries reflecting on the method itself: its strengths, its limits, its ethical implications, and its potential for education and training.
• Visual and infographic materials illustrating information flows and data metamorphosis across care chain actors.

How to participate

You can participate in several ways:

• Browse and use: all deposits are openly accessible and most are published under Creative Commons licences. You are welcome to read, cite, adapt, and build upon them.
• Submit your own work: if you have examples of AI-augmented reasoning applied to healthcare (or health-adjacent domains), you are encouraged to submit them. We accept working papers, annotated prompt logs, methodological reflections, commentaries, datasets, and software tools. Submissions go through a light curation review (see the Curation Policy page).
• Comment and discuss: Zenodo does not yet have a built-in discussion feature, but we welcome dialogue through linked platforms (LinkedIn, ResearchGate, email). Each deposit includes contact information for its authors.
• Propose collaborations: if you would like to co-develop a worked example or a methodological paper, get in touch.

Who we are

This community was initiated by Angelo Rossi Mori, a health informatician with decades of experience in clinical information standards, integrated care models, and healthcare terminology. The augmented reasoning approach emerged from his daily practice of using AI tools as reasoning partners for complex healthcare planning tasks — and from the conviction that this way of working deserves to be shared, examined, and improved collectively.

Contact: angelo.rossimori@cnr.it