Hollywood’s latest controversy is Tilly Norwood—an AI-generated “actress” that has talent agencies intrigued and SAG-AFTRA furious. The union’s message is clear: creativity must stay human-centric, and synthetic performers threaten jobs and authenticity. The fuss makes headlines, but it also highlights an issue in healthcare. As we face a rare shortage of clinicians, we should ask a question the film industry is grappling with: what role should AI play—and which roles must remain human?
The good: AI as the great understudy (not the star)
There’s a real, solvable shortage problem in care. We don’t have enough clinicians, especially in behavioral health, primary care, and specialties for smaller hospitals. Used well, AI can fill empty chairs without replacing the people we can’t afford to lose Hippocratic AI keeps its agents within non-diagnostic, protocol-driven roles—preoperative preparation, post-discharge follow-ups, and high-volume education—so nurses and physicians can concentrate on the ambiguous, the emergent, and the human.
Ellipsis Health considers the voice as a vital sign, signaling increased risk for depression and anxiety during routine interactions so clinicians can intervene earlier.
AmplifyMD expands specialists into “care deserts,” connecting limited expertise with the appropriate patients at the right time.
When we view AI as an amplifier, rather than an impersonator, access improves, triage speeds up, avoidable transfers decrease, and clinician burnout lessens.
The bad: unpredictable empathy and the trust cliff
Patients will forgive a clumsy app; they won’t forgive a synthetic “clinician” that sounds
compassionate but gets the chest pain playbook wrong. That’s the Tilly lesson for healthcare: if AI performs like humans instead of serving humans, trust collapses. The antidote is simple: no bots pretending to be licensed professionals; full transparency about who (or what) you’re talking to; and strict boundaries on scope with clear escalation to a human at the first sign of uncertainty.
The bridge across the cliff: create a trust layer that patients manage
Trust isn’t just a vibe; it’s the foundation. That’s why I’m optimistic about patient-controlled data and consent tools like HealthEx—a way to make provenance and permission understandable by computers. Imagine a world where patients create their records once, see a record of who accessed what and why, and can revoke permissions as easily as they granted them. Now, when Hippocratic’s agent calls to reconcile medications, or Ellipsis highlights risks, or an AmplifyMD specialist opens a chart, the patient has authorized that data flow—and can review it. That’s how you prevent the Tilly problem in a clinic: no impersonation, no mystery, full traceability.
The missing curriculum: Humanics (teaching humans to be better humans than AI)
Here’s the part we never discuss openly: we still admit future doctors using a centuries-old filter—science GPA, MCAT scores, and whether you had a love affair with organic chemistry—and then we’re surprised that some graduates struggle with empathy, creativity, and communication. Meanwhile, AI models are becoming frighteningly proficient at the very skills we value: memorization, pattern recognition, and recall. If we don’t change who we admit and what we teach, we’ll continue to graduate brilliant diagnosticians who are unprepared for the one job machines can’t do: be human on purpose.
Enter Humanics, a clear, revolutionary framework: technological literacy, data literacy, and human literacy. Every medical and nursing student needs foundational skills in the tools of the fourth industrial revolution, the ability to interpret the data these tools produce, and, most importantly, the development of distinctly human qualities. It even features a motto crafted for this moment: “AI won’t steal your job, but someone who works with AI will!”
Human literacy isn’t vague. It’s creativity, critical thinking, entrepreneurship, and cultural agility, the very skills our current admissions processes and curricula undervalue, and that AI can’t truly replicate. Yes, we still primarily select the next generation of physicians based on science GPA, MCAT scores, and organic chemistry grades. Yet, we are surprised that some doctors lack empathy, communication, and creativity.
So, what does “teaching humans to be better humans than AI” look like in practice?
Admissions 2.0: Improve value as much as immunology. Use structured behavioral interviews, team-based design challenges, and assessed community engagement alongside MCATs.
Curriculum 2.0: Shift from “don’t use AI” to “use AI responsibly,” and evaluate the important human work products like explaining uncertainty, co-creating care plans, and navigating cultural nuances.
Assessment 2.0: OSCEs that evaluate listening, curiosity, and co-regulation, not just checklist empathy.
Proof of concept: the Fairleigh Dickinson University model
At Fairleigh Dickinson University, we’re embodying Humanics through the Humanics Innovation Challenge Fund—a multidisciplinary effort to develop “humanAI translators” that merge intellect, emotion, and algorithms. The fund supports four labs that represent the future I want my own family to adopt, including a Healthcare Humanity Lab, CultureTech Collision Studio, a Storytelling Data Lab, and a Creative Disruption Incubator.
Students pitch “impossible partnerships” in two-minute videos, prototype with grants, and showcase their work at a community symposium; the early fund commits $200,000 with room to grow. Translation: we’re not teaching around AI; we’re teaching through AI to elevate human work.
Final scene
Tilly Norwood can headline a film festival panel, but she can’t scrub in. If we develop a patient-controlled trust layer and reframe medical education around Humanics—integrating technology, data, and human literacy—we won’t fear AI; we’ll leverage it to make clinicians more compassionate at critical moments. Remove the algorithm as our gatekeeper in organic chemistry. Instead, employ it as a co-pilot. Then, admit and train doctors to focus on the only scalable competitive advantage in the AI era: being brilliantly, courageously human.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
