TLDR: Google has presented two studies demonstrating how generative AI tools, including its LearnLM model, can significantly improve medical education by supporting clinical reasoning skills and offering scalable learning alternatives. These findings, shared at the MedEd on the Edge conference, come as the global health workforce faces a projected shortage of over 11 million professionals by 2030.
As the global health workforce anticipates a projected shortage of over 11 million professionals by 2030, Google is actively exploring the transformative role of Artificial Intelligence (AI) in reshaping medical education. The tech giant recently presented two comprehensive studies at the MedEd on the Edge conference, underscoring AI’s potential to bolster healthcare training for both students and educators.
The first study, titled ‘Generative AI for Medical Education: Insights from a Case Study with Medical Students and an AI Tutor for Clinical Reasoning,’ was published at CHI 2025. This research delved into how medical learners interacted with an AI tutor prototype. Through a series of workshops, prototyping, and user studies, researchers observed a strong interest among participants in AI tools capable of adapting to individual learning needs, providing constructive feedback, and fostering critical thinking skills.
The second study focused on LearnLM, Google’s Gemini-based family of models specifically fine-tuned for educational purposes. This research involved medical students and physician educators, comparing LearnLM’s performance against a base model across 50 synthetic evaluation scenarios that covered both preclinical and clinical topics. The results were compelling: across 290 conversations, physician educators consistently favored LearnLM, noting its superior ability ‘to demonstrate pedagogy and behave more like a skilled human tutor.’ Medical students also rated LearnLM more positively in terms of overall enjoyment and interaction.
Collectively, both studies suggest that generative AI tools can play a crucial role in developing clinical reasoning skills, while simultaneously offering scalable alternatives to address existing gaps in health education. Google emphasized that these evaluations were conducted with a strong focus on responsible implementation and human oversight, ensuring no patient data was utilized in the studies.
Also Read:
- University of Copenhagen Research Reveals AI Tool Outperforms ChatGPT in Learner Support
- Google Australia Launches Campaign Showcasing AI’s Real-World Impact in Healthcare
Looking ahead, Google highlighted the imperative to develop curricula that not only integrate AI as a practical tool but also educate future healthcare professionals on how to understand and responsibly utilize this technology. The company plans to continue its collaboration with educators to design personalized learning experiences, further strengthening medical education through the integration of LearnLM capabilities into Gemini 2.5 Pro.
