GTC DC 2025 Agentic AI + Patient-First Digital Health

Category: Healthcare Special Address · Year: 2025 · ▶ Watch

Switch language → 中文

Segments (14)

  • 00:00 · Introduction
    • Trent Norris introduces the panel and asks why now is the time for AI in healthcare.
  • 00:33 · The Transformational Opportunity of AI
    • Stephen Gillett discusses how AI will touch every industry and person, marking a transformational moment.
  • 01:43 · Structural Challenges in Healthcare
    • Stephanie Carlton highlights lagging productivity and rising costs in healthcare, and how AI can address clinician burnout.
  • 04:18 · Data Foundation and AI Impact at Mayo Clinic
    • Kent Thielen explains how Mayo Clinic leverages its vast digital data, including digital pathology, to improve diagnostics and care.
  • 06:59 · Addressing the Clinician Burnout Crisis
    • Shiv Rao discusses the public health emergency of clinician burnout and how agentic AI can automate clerical tasks.
  • 09:30 · The Patient-Clinician Experience
    • The panel discusses how tools like Abridge are changing the interaction between patients and clinicians.
  • 12:10 · Changing Patient Expectations
    • Kent Thielen notes that patients expect more digital capabilities and shares a story of a clinician saving significant time using AI.
  • 15:33 · Data Curation and the Verily Me App
    • Stephen Gillett talks about the challenge of curating valuable data from noise and introduces the Verily Me app.
  • 18:27 · CMS Initiatives for Data Interoperability
    • Stephanie Carlton outlines CMS’s goals to make health data accessible, interoperable, and usable for patients.
  • 22:40 · Building Trust in AI
    • Shiv Rao emphasizes the importance of transparency, reliability, and credibility in building trust for AI in healthcare.
  • 25:42 · Partnerships and Trust at Mayo Clinic
    • Kent Thielen discusses how Mayo Clinic evaluates partnerships and tools to ensure they are reliable and user-friendly.
  • 28:20 · Trust and Safety in Corporate Culture
    • Stephen Gillett stresses that trust and safety must be foundational to a company’s culture when developing AI.
  • 33:39 · Balancing Innovation and Safety at CMS
    • Stephanie Carlton discusses how CMS aims to promote innovation while maintaining safety and patient consent.
  • 35:38 · Conclusion
    • Trent Norris concludes the panel with a prediction about the future of agentic and physical AI.

Product Announcements (1)

  • [16:34] Verily Me
    • An app that connects patients with clinicians and brings their health data to bear in that relationship.
    • specs: Aggregates historical health data (e.g., procedures, vaccines, doctor visits) to make it accessible for clinical encounters.
    • availability: Launched recently (as of the recording).

Specific Numbers (10)

Timestamp Metric Value Context
01:57 Productivity lag 4 to 10 times Healthcare productivity lags behind other industries by this factor.
05:22 Digitized slides 17 million Number of pathology slides digitized at Mayo Clinic.
06:44 Time translation 1 hour to 7 hours One hour of developer time translates to seven hours of autonomous time due to agentic AI.
07:16 Doctor burnout 2 out of 5 Doctors who don’t want to be doctors in the next 2-3 years.
07:20 Nurse burnout 27% Nurses who don’t want to be nurses in the next 12 months.
10:01 Tool access 50% Clinicians worldwide who have access to tools like Abridge.
15:08 Time saved 2+ hours a day Time saved by a clinician using Abridge for clinical notes.
16:54 Health history 23 years Amount of personal health history Stephen Gillett has in his app.
20:23 Committed companies 60 Tech and tech-forward companies committed to CMS’s interoperability framework.
27:39 Platform lives 15 million Number of lives on Mayo Clinic’s curated data platform.

Customer Stories (1)

  • [14:24] Mayo Clinic Clinician (Kent Thielen’s wife)
    • Adopted the Abridge tool for clinical documentation.
    • outcome: Saved 2+ hours a day on clinical notes, allowing more time for teaching and research, described as the most transformational thing in 30 years of practice.

Key Technologies (4)

  • Ambient AI: Listens to patient-clinician interactions to generate structured notes.
  • Digital Pathology: Digitizes physical biopsy slides for analysis and historical reference.
  • Agentic AI: AI that can autonomously orchestrate tasks and reason through problems.
  • Large Language Models (LLMs): Generates text and answers questions based on prompts.

Predictions / Commitments (2)

  • [10:28, Next 5 to 10 years] We will not be able to fully automate a doctor or a nurse.
  • [36:36, Next couple of years] Agentic AI will collide with embodied AI/physical AI to augment the workforce.

Companies Mentioned (5)

Abridge · Garmin · OpenAI (ChatGPT) · Alphabet/Google · Netflix / Amazon

Notable Quotes (3)

We’ve been using software as picks and shovels or tools, and now that software is using the tools. — Trent Norris @ 04:02

Clinicians, we don’t get compensated for the care that we deliver, we get compensated for the care that we documented that we delivered. — Shiv Rao @ 10:56

If trust breaks down, everything breaks down. — Kent Thielen, M.D. @ 26:21

Key Topics

Agentic AI · Healthcare Productivity · Clinician Burnout · Ambient AI · Digital Pathology · Data Interoperability · Trust and Safety in AI · Patient Experience · CMS Regulations · Healthcare Partnerships · Data Curation · Value-Based Care

Takeaways

  • AI is transitioning from being a passive tool to an active, autonomous agent in healthcare.
  • Clinician burnout is a severe crisis; AI tools like ambient scribes are crucial for reducing administrative burden.
  • Healthcare data is abundant but historically siloed; efforts are focused on curating it and making it accessible to patients.
  • CMS is pushing for greater data interoperability and patient access to their longitudinal health records.
  • Trust, transparency, and reliability are the most critical factors for the successful adoption of AI in clinical settings.
  • Patient expectations are evolving, demanding consumer-like, data-driven experiences similar to other tech platforms.
  • Corporate culture must prioritize safety and trust when developing and deploying AI technologies in healthcare.