Welcome back to The AI Pulse Weekly, your curated digest on the convergence of artificial intelligence and healthcare.
This week, the narrative of AI in medicine is one of powerful dualities. On one hand, we see innovation pushing the boundaries of what's possible, from developing autonomous medical assistants for deep space missions to designing diagnostic tools that learn with unprecedented efficiency. These advancements promise a future of more personalized, predictive, and accessible care.
On the other hand, this wave of progress is met with critical, grounding realities. The very infrastructure of healthcare data remains a significant bottleneck, threatening to stall the deployment of these advanced tools before they can achieve scale. Simultaneously, high-profile failures, such as AI models "hallucinating" medical facts, serve as stark reminders of the profound risks involved. This tension underscores a pivotal moment for the industry, where the path forward requires not only brilliant engineering but also a deep commitment to building robust data foundations, establishing rigorous ethical safeguards, and fundamentally rethinking the roles of human experts in a world of augmented intelligence.
And then to cap it all off, GPT-5 dropped, which took steps to improving its HealthBench performance and substantially reduce hallucinations.
I am also thrilled to host Enoch Kan, CTO at Root Note Ventures who advises on AI and Health tech, as a guest author to share the zeitgeist of agentic AI in the clinical environment.
Let's dive into the headlines making waves this week…
Your Weekly Dose of AI in Health
OpenAI CEO Sam Altman Says GPT-5 Should Be Used for Health: 🩺 OpenAI has released GPT-5, its most advanced reasoning model, with a significant focus on healthcare applications. CEO Sam Altman emphasized that health is a primary use case for ChatGPT, and this new version aims to empower users in their healthcare journey by helping them understand test results and formulate questions for their doctors.
NASA and Google Partner to Develop AI Medical Assistant for Astronauts on Mars 🧑🚀: In an ambitious collaboration, NASA and Google are creating an AI-powered medical assistant to support the health and well-being of astronauts during long-duration missions to Mars, providing autonomous diagnostic and treatment recommendations in deep space.
Report Highlights Instances of Google's Medical AI 'Hallucinating' Inaccurate Brain Information 🧠: A recent report indicates that Google's medical AI has, in some instances, generated incorrect information about brain anatomy, underscoring the critical need for rigorous validation and oversight in AI applications for healthcare.
Gates Foundation Pledges $2.5B Towards Women's Health Initiatives, Including AI in Low-Income Nations 💰: The Bill & Melinda Gates Foundation has committed a substantial $2.5 billion to bolster women's health globally, with a significant portion dedicated to leveraging AI technologies to improve healthcare access and outcomes in low-income countries.
Industry Leaders Emphasize Critical Need for Healthcare Data Interoperability 🔗: Key stakeholders in the healthcare sector are reiterating the urgent requirement for seamless data exchange between different systems to unlock the full potential of AI-driven diagnostics and personalized medicine, ultimately improving patient care coordination.
Each week, we select a critical topic for an in-depth exploration.
Forget the Chatbot: Healthcare's Future is a Symphony of Specialized AI Agents
By Enoch Kan, Root Note Ventures CTO
Healthcare is about to change — and it won’t be because of a chatbot.
At Hackensack Meridian Health (HMH), one of New Jersey’s largest health systems, they’re piloting an agentic AI platform with an orchestrator managing multiple specialized agents. Booking follow-up appointments, verifying insurance, coordinating discharge tasks — all handled seamlessly.
And they are not alone.
We’re seeing a quiet revolution where large healthcare providers, tech giants, and startups are rethinking how we interact with care. Instead of a single AI assistant trying to do everything, we’re entering the era of AI agent orchestration — where many agents, each with a specific function, work together behind the scenes.
Let’s break it down.
What’s actually happening?
1. Multi-agent orchestration is going mainstream
Microsoft is already running clinical pilots where diagnostic agents “debate” each other to reach the most likely diagnosis. Their Microsoft AI Diagnostic Orchestrator (MAI-DxO) outperforms human doctors on tough clinical cases, while making its reasoning transparent.
Commure just launched “Commure Agents,” plug-and-play agents embedded in EHR workflows to handle scheduling, billing, and even discharge planning.
And Salesforce is rolling out Agentforce for Health, with libraries of agents automating everything from patient intake to trial matching — all with built-in auditability.
2. The assistant is no longer just a chatbot
We’ve moved beyond “Hey Siri for healthcare.” These assistants don’t just respond — they route, infer, and collaborate.
Hackensack Meridian’s pilot agentic assistant, for example, might delegate:
an appointment to a scheduling agent,
a prescription refill to a pharmacy agent,
a coverage check to a benefits agent,
and loop in a clinical agent for context — all without you repeating yourself.
3. AI agents are learning to reason (not just regurgitate)
The best orchestrators today use tool-augmented agents. That means each one has access to databases, APIs, or even search tools to verify or supplement its answers. In Microsoft’s MAI-DxO, diagnostic agents use real clinical tools and reference materials to back up decisions.
Academic projects like MedOrch and MATEC (Multi-AI Agent Team Care) show agents collaborating in sepsis care or Alzheimer’s diagnosis, using modular toolkits with explainability built in.
4. Trust and transparency are non-negotiable
The stakes are high. These systems deal with clinical risk and sensitive patient data. So companies and healthcare providers like Hackensack Meridian Health, Google Cloud, Microsoft, and Commure are building human-in-the-loop, privacy-preserving, and audit-friendly frameworks by design. In other words: if an AI recommends something, it better be traceable, reversible, and verified by physicians.
Why this matters now
The global digital health market is exploding — $500B+ by 2025. But the winners won’t be those who just slap a chatbot on top of an app. They’ll be the ones who:
unify backend workflows,
integrate agents directly into clinical and operational systems,
and build trust across every user touchpoint — from patient to provider to admin.
Think about the possibilities:
For patients: smoother experiences, having access to unified health records without the frustration of bouncing between systems.
For clinicians: less administrative overload and better diagnostic tools.
For health systems: new revenue models from AI-driven insights, virtual care coordination, and continuous patient engagement.
Final thought
The future isn’t one assistant that knows everything. It’s many agents that know exactly what they’re good at, working together in real time, orchestrated with context, governed with care.
HMH is going live. Microsoft already is. Commure is productizing. The (multi-)agentic era of healthcare has arrived — and it’s not about building a smarter, all-encompassing chatbot. It’s about building a smarter system.
Are we ready?
Stay informed on frontier research on the future of AI and health.
Study Reveals AI Chatbots Susceptible to Medical Misinformation 🛡️: Research indicates that current AI chatbots can sometimes generate inaccurate or misleading medical information, highlighting the crucial need for robust safeguards and careful validation before widespread clinical deployment.
AI Shows Promise in Early Prediction and Prevention of Heart Disease in New Study 🩺: A new study demonstrates the potential of AI algorithms to analyze patient data and identify individuals at high risk of developing heart disease, paving the way for proactive and preventative healthcare strategies.
NEH Awards Grant to Launch Center for Humanities-Based Health AI Innovation at Baylor and Rice 💡: The National Endowment for the Humanities has awarded a grant to Baylor College of Medicine and Rice University to establish a center dedicated to exploring the ethical, social, and humanistic implications of AI in healthcare innovation.
Novel AI Tool Learns to Interpret Medical Images with Significantly Less Training Data 📸: Researchers have developed a groundbreaking AI tool that can effectively learn to interpret complex medical images using a fraction of the data typically required, potentially accelerating the development of AI diagnostics.
FIU Researchers Advance AI for Enhanced Heart Disease Prediction and Prevention 📊: Continuing their work in cardiovascular health, FIU researchers are making strides in leveraging AI algorithms to improve the accuracy and efficiency of heart disease prediction and develop targeted prevention strategies for at-risk individuals.
Mark your calendars for essential industry gatherings and educational opportunities.
Event | Date | Sponsor |
|---|---|---|
October, 10, 2025 1 p.m. – 4 p.m. San Diego, CA | American Medical Association | |
October 19-21, 2025 Pittsburgh, Pennsylvania | The University of Pittsburgh |
Reach out if you have an event you’d like to promote [email protected]
Help us grow our community and spread the word about the exciting advancements in AI and Health.
If you found this newsletter valuable, please share it with your network with this Referral Link!
Thank you for reading The AI Pulse Weekly. I’d love your feedback, so please drop me a note at [email protected] with thoughts, suggestions or feedback. I'll be back in your inbox next Friday with more Health and AI insights!
Have a great week ahead!
Sean
Copyright © Root Note Ventures LLC, All rights reserved.