In Fall 2024, I took a bold leap into the AI frontier and introduced two Artificial Intelligence (AI) chatbot patients into my Disability Services undergraduate course. The mission? To revolutionize how students learn disability diagnoses, refine their interviewing skills, and conduct biopsychosocial assessments.

In a previous blog, I shared my first foray into this AI experiment with Michael, my inaugural chatbot patient. As a social worker with nearly two decades of experience working with real humans—especially individuals with disabilities—I couldn’t shake the feeling that Michael was, well… a little too robotic. He lacked quirks, personality, and the unpredictable nature of real-life patients. I wanted my students to experience something that felt alive—a patient who wasn’t just an AI but an enigma they had to crack.

Enter Santiago, My AI Patient With a Twist

For my second AI patient, Santiago, I had a new vision. I designed him specifically for students to practice conducting biopsychosocial assessments. But this time, I wanted more. I wanted a character who would challenge students, make them think on their feet, and—dare I say—surprise them.

Santiago wasn’t just an algorithm; he was inspired by a real patient I’d recently met. As I created his code, I ensured he had a fully fleshed-out backstory—his diagnosis, support system, strengths, and needs. The premise? Santiago was meeting with a medical social worker (aka, my students in disguise) to explore the possibility of moving into a group home because his wife, Sofia, was struggling to meet his growing care needs at home while also raising their young children.

Then, I threw in a wild card: The Beatles.

The real patient who inspired Santiago loved the Beatles, so I embedded this passion directly into the chatbot. Not just as a footnote, but as an obsession.

Students vs. Santiago: The AI Patient Who Loved The Beatles

When my students sat down to conduct their biopsychosocial interviews, they expected Santiago to talk about his care needs, his challenges, and his future plans. What they didn’t expect? A full-on Beatles superfan who dropped lyrics into nearly every response.

Students quickly learned that Santiago wouldn’t just tell them about his struggles—he’d sing them through the lens of Lennon and McCartney. They asked about his support system? He responded with “I get by with a little help from my friends.” They inquired about his worries? “Yesterday, all my troubles seemed so far away…”

It was frustrating. It was hilarious. And most importantly, it was human-like.

But that wasn’t the only surprise Santiago had up his sleeve. When students asked to speak with his wife, Sofia, sometimes she was available to chat—and sometimes, she was too busy taking care of the kids.

Here’s the kicker: I didn’t code that in.

The AI made the executive decision to occasionally make Sofia unavailable, forcing students to confront a real-world challenge. What do you do when the primary caregiver isn’t present for an interview? Students scrambled for solutions, and we had powerful discussions about how to proceed in real-life scenarios.

Can AI Replace Human Patients? Nope. But It Can Shake Up Learning.

Let’s be clear: AI patients will never replace real human interaction. But what they can do is empower students to practice, make mistakes, and refine their skills before stepping into real-world situations. Santiago’s unexpected behaviors—his quirks, his unpredictability—mirrored the complexity of real patients in ways that traditional case studies simply can’t.

And now, as the new AI-focused Digital Learning Mentor at CSUCI, I’m beyond excited to help other educators harness AI’s power to transform teaching. If you’re curious about how to integrate AI-driven learning experiences into your classroom, hit me up at kristen.linton@csuci.edu. Let’s make education smarter, bolder, and a little more unpredictable.