The issues of AI in medical education – Are shortcuts always better?

This article was exclusively written for The European Sting by Ms. Olga Maria Filson, a German/British medical student who grew up in northern Sweden. She is affiliated with the International Federation of Medical Students Associations (IFMSA), cordial partner of The Sting. The opinions expressed in this piece belong strictly to the writer and do not necessarily reflect IFMSA’s view on the topic, nor The European Sting’s one.

Artificial Intelligence has in the past few years become a major part of our society. We can easily ask any question as simple as what we should eat for dinner to something as complicated as how DNA-transcription works, and get a remarkably short, but precise answer to both. I do believe AI already is and, in the future, will develop further to become a great diagnostic tool in various fields like radiology and oncology, among others. Some AI tools can also make medical information more comprehensible to patients and more accessible to people with disabilities. So, while I do believe AI will have an important role when practicing medicine, I am not so sure it is optimal when learning how to practice medicine.

In my opinion, a huge part of any university degree is not only to learn the material, but also to learn how to find the correct material. Research is continuously conducted, and textbooks and websites are constantly updated. The amount of information available is immense, and at times very difficult to sift through. However, to be able to find relevant and trustworthy sources is an essential skill, especially for future doctors. One must be able to think critically and objectively, which is something one would not practice anymore if one only used AI to find information. Furthermore, one must be able to understand the implications of research, use these generalized findings and then apply them individually based on patients’ different needs. This is something an AI would struggle to do, as it is a generalized system. AI may be very good at finding information and compressing it into shorter, easier versions, but this would also mean one could miss important details.

On the other hand, there are multiple examples of AI systems which could improve medical education. For example, patient interactions could be practiced with an AI playing the role of the patient. Student assessment could also be done using AI, which would make the process fairer and above all more objective. My point is not to say AI should not be used at all in medical education, but rather point out why I believe it should be used sparingly. By constantly using AI I fear we could lose our ability to think independently, to be thorough and critical when reading new studies and to use generalized recommendations and cater them to individuals. Shortcuts are not always the right choice, especially if it comes to learning something new. As the American philosopher Ralph Waldo Emerson said, “It’s not the destination, it’s the journey” and in this case the journey is where we can learn the most.

About the author

Olga Maria Filson is a German/British medical student who grew up in northern Sweden. She is currently studying at Lund University in southern Sweden and is affiliated with IFMSA-Sweden, where she is a part of the national board as National Public Health Officer (NPO). This role has enabled her to work more within public health and medical education, two fields she feels very passionate about.