Faces of Digital Health

Most conversations about agentic AI in healthcare get stuck on capability. This one is about the gap between capability and deployment — and what closes it.

Aashima Gupta, Global Director of Healthcare Strategy and Solutions at Google Cloud, argues that healthcare's bottleneck isn't vision; it's courage. The processes are documented poorly or not at all, AI fluency programs reach a fraction of employees who want them, and most enterprises are running agents without the harnesses — grounding, evaluation, red-teaming — that production deployment actually requires. Meanwhile patients navigate three different "clock speeds" (annual insurance cycles, shifting provider rosters, Medicare pricing) that bear no relation to the timeline of their own health.

We cover the European vs US deployment posture, the difference between agents-with-agency and rule-based AI, why Highmark's library of one million internal prompts matters, Google Cloud's full-stack efficiency play (TPU Ironwood, Gemini, the 40% data-centre electricity reduction DeepMind delivered years ago), and the multi-agent "harnesses" — including the red/blue/green team architecture — that are starting to make production-grade healthcare AI plausible.

Video: https://youtu.be/rLtaxQLgCg0?si=JDP6kK97_tYsFoSb

Newsletter: https://fodh.substack.com/

Agentic Patient Series: https://www.facesofdigitalhealth.com/agentic-patient-blog

The Agentic Patient is here — and most healthcare systems don't have a plan for it. In this special reverse-role episode of Faces of Digital Health, Eric Sutherland interviews host Tjaša Zajc about what a year of using AI through her own chronic illness has actually taught her about patients, doctors, and the future of healthcare AI.

200 million people will ask ChatGPT a health question this week. The question is no longer whether patients will use AI to navigate their care — it's how to help them do it well, without harm, and in productive partnership with their clinicians.

In this episode:
- Why "patients know best" breaks down for chronic patients
- The three archetypes AI is creating: minimizers, cyberchondriacs, and informed collaborators
- What happens when doctors dismiss patients who use AI
- A two-model verification method for cross-checking medical AI advice
- Why "digital literacy" is the wrong name for the most important skill in modern healthcare
- Two prompts that genuinely change what AI gives you back
- What health ministries should actually do — and why we shouldn't offload patient AI education to doctors

⏱ CHAPTERS
00:00 Intro & reverse-role experiment
01:00 Eric Sutherland: "a data guy with personality"
01:36 A year as a chronic patient using AI
02:50 Same prompt, different LLMs — the trust problem
04:30 How The Agentic Patient series was born
06:00 Three patient archetypes
09:00 When doctors dismiss AI, patients start hiding
12:30 Dale Atkinson, HIMSS Europe, and data outside the clinic
13:30 200M weekly ChatGPT health queries — who's accountable?
15:30 The two-model cross-verification method
17:00 Making 7-minute appointments work with AI
19:30 Finland's Elements of AI — a model for healthcare
22:00 Why chronic patients may not know best
24:30 Five minutes with a health minister
27:00 Two prompts that change AI outputs
30:00 The agentic patient is a survivor, not a tech enthusiast

🎙 ABOUT THE AGENTIC PATIENT
The Agentic Patient is a series under Faces of Digital Health exploring how patients and clinicians are actually using AI in healthcare — the wins, the harms, and the best practices emerging across cancer care, chronic disease, and primary care.

🔗 LINKS
Newsletter: https://fodh.substack.com/p/the-agentic-patients-are-here
More episodes: https://www.facesofdigitalhealth.com/agentic-patient-blog
Tjaša Zajc on LinkedIn: https://www.linkedin.com/in/tjasazajc/
Eric Sutherland on LinkedIn: https://www.linkedin.com/in/esutherland272/?skipRedirect=true

#AgenticPatient #AIinHealthcare #DigitalHealth #FacesOfDigitalHealth #HealthcareAI #ChatGPT #PatientEmpowerment #ChronicIllness #AIliteracy #MedicalAI #PatientAdvocacy #DigitalTransformation

When Demetri Giannikopoulos was diagnosed with multiple sclerosis, his community neurologist handed him a sheet with fifty medication options and told him to pick one. That was a long time ago. Today he's the Chief Innovation Officer at RadAI, overseeing how artificial intelligence gets deployed in radiology across US health systems — and he's spent two decades learning how to navigate a healthcare apparatus that, in his words, "is not designed for sick patients."

In this conversation Demetri explains why the most valuable thing AI has done for him as a patient isn't clinical — it's the 50 pages of insurance underwriting documents he fed into ChatGPT to save several thousand dollars on a plan that looked, on paper, worse. He walks through his "red team" prompting technique, the error he caught in a radiology report where legacy speech-recognition software had dropped the word "no," and why he thinks the regulatory debate around AI in healthcare should look less like drug approval and more like how we regulate nuclear power. If you want a ground-level view of what AI can and cannot do inside the American medical system, this is where to start.

Additional resource with prompt tips: https://aipatients.org/
Additional resource: Scanxiety toolkit: https://edge.sitecorecloud.io/americancoldf5f-acrorgf92a-productioncb02-3650/media/ACR/Files/Clinical/Patient-Family-Centered-Care/PFCC-Scanxiety-Toolkit-Brochure-Digital-Version.pdf Full Agentic Patient series: https://www.facesofdigitalhealth.com/agentic-patient-blog

Detailed summary and tips from Demetri: https://www.facesofdigitalhealth.com/agentic-patient-blog/red-teaming-your-health-plan-demetri-giannikopoulos-on-responsible-ai-the-cures-act-and-what-patients-should-actually-do

6 tips on AI use for patients: https://fodh.substack.com/p/the-agentic-patients-are-here

Diana Ferro works at a major pediatric hospital in Italy, working on AI infrastructure, rare diseases, and — importantly — the International Alliance of Pediatric Centers on AI. Unlike the patient voices earlier in the Agentic Patient series, she sits on the other side of the consulting-room door. Her concerns are sharper, more specific, and more uncomfortable. She is not against patient AI use. She is watching what happens when desperate parents, teenagers in crisis, and sycophantic chatbots meet in a pediatric setting and she is trying to build the guardrails in real time.

Diana frames AI in pediatric medicine as a two-front problem. On one front, Italian hospitals are racing to build the data infrastructure — EU-funded — to share research across institutions and turn billing data into diagnostic and predictive tools. On the other front, patients and families are already ahead of the system, using consumer LLMs in ways that clinicians are not trained to respond to.

She describes three specific, observed harms she's seeing in pediatric practice:


parents using AI to deny rare-disease diagnoses,


adolescents using AI as a pro-eating-disorder coach by telling it they want to "lose weight to be healthy,"


young people with weak support systems finding AI easier to talk to than a clinician — including, she notes, in contexts tied to self-harm.

The Agentic Patient Series: https://www.facesofdigitalhealth.com/agentic-patient-blog

Agentic Patient 6 tips: https://fodh.substack.com/p/the-agentic-patients-are-here

Russ was diagnosed with bowel cancer in late 2021 and simultaneously with smoldering myeloma, aged 40. The smoldering myeloma has been inactive; the bowel cancer has progressed through multiple surgeries (bowel, liver, lung) and is now stage 4, on active chemotherapy. He runs AI for the business he works for, so his day job is adjacent to the technology. He blogs publicly about his disease at fcancerwith.ai and on LinkedIn. He is British; cared for by the NHS with some private care around the edges. He is articulate, technically fluent, and willing to pay roughly £200 a month for AI subscriptions.