AI and Continuous Medical Education: How Far Can It Go?

When I first became a dad, I was told days would run slow but years could happen in the blink of an eye. That phrase had never made more sense than now, when arguably the most famous toddler in the world has turned 2 years and 4 months old: ChatGPT. Currently, almost every active professional uses AI daily (whether by choice or not), but could you imagine that ChatGPT did not exist before November 2022?

While there’s ongoing hype about the potential of artificial intelligence, little is known about how it could be applied properly in education, let alone medical education. When you think about it, it makes a lot of sense: education hasn’t changed dramatically in the last few decades, nor has continuous medical education. So, where do we stand in adapting new technologies in the pharmaceutical and healthcare industries, and how could we benefit from them?

Roy Amara said it best: “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” While there’s a rush to apply AI to every single aspect of our daily professional routine, we tend to get distracted and miss the big picture. Don’t get me wrong, applying this technology to e-learning platforms by generating videos with AI and displaying them with interactive avatars is a great way to maximize resources, but more than reducing production costs, does it really make an impact on the way physicians are learning? What could we do better, and which path should we take?

The good news is we are not alone here, according to a 2024 study from the journal “Medicina Clínica Práctica” in Barcelona, “It is quite clear that the lack of effective integration of artificial intelligence in medical education prevents access to the benefits of its use in medicine and may even have adverse effects on patient care.” 

I understand there are several opportunities and use cases for AI in medical education (such as virtual reality, clinical simulation, image analysis, and generative AI); however, I’d like to focus on the one I believe could get the most out of this technology: Hyper individualized learning.

Imagine starting a course about ophthalmology that has no pre-defined modules, that starts with no syllabus, and has no defined tests. That’s it, as soon as you sign up, you start a conversation with an AI-instructor who has complete interest in you, in your objectives, in the time you have available, and wants to know what type of learning works best for you. Just a quick chat with someone (something) who wants to design the perfect learning strategy for you and the needs you have whenever you start this learning process. Now that’s personalization.

These types of learning methods could have benefits like:

• A Complete Customized Course: Every learner would have a unique experience tailored to their specific needs and preferences. No two individuals would follow the same learning path or receive identical content.

• Self-guided learning / Self-paced: The course would adapt to each learner's pace and style, allowing them to learn at their own speed and on their own terms.

• Dynamic Adjustments: Whenever a user needs guidance or a change of direction, the course would learn from their previous interactions and adjust the content, pace, or style properly.

• Continuous Feedback: Every user would receive consistent and personalized feedback to help them monitor their progress and identify opportunity areas for improvement.

• Adaptative Gamification: Every user would be available to adapt the amount of gamified experience in order to gain the most out of the content.

This is one of the many approaches that could apply artificial intelligence to the daily practice of continuous medical education, and even when it needs to be polished, it can be a great starting point for further implementations. It does seem like a bold move to integrate such an innovative step into a company’s strategy, especially in an industry like healthcare where regulations can hardly keep up with technology; but in a professional environment where everyone is trying to differentiate, it is sure worth the risk. 

Introducing new technologies can always be scary, and it does not always work the first time. It comes with challenges: from user adoption to lack of experience in implementation; but once it is efficiently applied, it seems like the obvious choice looking back. 

Returning to my first analogy, whenever your son or daughter falls trying to take their first steps, you don’t say, “Walking is not for you.” I’m sure you tell them to try again.