When I was a kid, my Uncle Smiley was an expert in everything. No matter what question we asked him, he had the answer. He knew baseball, the solar system, the dinosaurs, the crops and how to ride a bike, to name a few. He was an amazing man and, of course, as kids, we never asked him to verify these “facts” that he would spew from memory.
I have gone to school for 26 years, 13 years from kindergarten to high school and 13 more years to get four degrees ending with a PhD. In those years, I also worked and taught, mostly in hospitals. So, my education, training and experience classifies me as an “expert” in legal cases. I have worked as an expert for more than 30 years and not once in that time did a lawyer or jury member ask me to verify the facts in a case or my opinions. However, I cannot make things up out of thin air, because there is an expert on the other side of the case to call me out if I did that. My expertise in legal matters is from years of study and work.
I am a professor at the University of Nebraska and my students will often ask me about how to manage an issue with a patient, and when the issue is complex, I just tell them what needs to be done. None of them have asked me where I got the information.
So, what am I leading to? Fairly often, practitioners will send notes or ask questions after lectures. For example, I was recently “told” that the only cause of nonblanchable erythema was grade 1 pressure ulcer. Of course, I explained that grade 1 pressure ulcer is one example of nonblanchable erythema, but not the only one. I went on to explain that chronic irritant dermatitis and other causes of purpura would not blanche. You know where this is going don’t you? The next question to me was “Is that published somewhere?” Well, I don’t know off hand – do you expect me to look it up for you?
What’s the difference? In these queries, which are sometimes to settle an argument, I was asked to serve as an expert. Obviously, my experience and education is not enough and everything must be backed up with published material. When I consider the errors in artificial intelligence (AI)-generated material, including making up references, how is having something published going to verify its accuracy? I realise that most professionals have been taught to “use the literature” as a basis for policy, procedure and practice. What then is the role of the expert? If you can look it up with AI, you will have an answer. I just cannot guarantee it will be accurate. Just because something is published, does not guarantee its accuracy. Consider the web as a common source of information, but it is not always accurate.
In case you are curious, I did look at the British Journal of Dermatology, which states: “Allergic contact dermatitis typically does not blanch when pressure is applied. This is due to the complex tissue changes involved in the condition, including edema and cellular infiltration, which are a result of a delayed hypersensitivity reaction (type IV) to an allergen that contacts the skin.” And AI says: “While acute rashes often blanch, chronic contact dermatitis does not typically blanch when pressed.”
So, what is the role of the expert? If everything needs to be published to be true, then will AI and AI-generated articles that are published in journals reduce the need for the expert? I certainly hope not. Experts are able to find the truth even when the literature leads in a different direction. I wonder what Uncle Smiley would say?
