Widespread hands-on access to AI is the latest fad, but it won't fade away. I'll be curating more posts on AI, and this one is my inaugural post. I am no expert on medicine but I will provide some (hopefully) insights and opinions on AI and its application within patient/provider interactions and what we may be seeing.
Since AI is accessible to everyone with a computer, we can begin to think about our patients coming in with questions and opinions about what they have, how they should pursue treatments, and even what their prognosis may be. Providers are already familiar with patients who Google and research their medical situations and must manage these expectations. I don't think this AI is going to be much different, with patients coming in with the knowledge they've been pointed to. Maybe the difference is that more patients will be bringing this information in because access will be easier for a wider range of people.
When I consider the provider's perspective, I see really helpful areas of summations and large data presentations that the provider can use in their guidance of testing and treatments. As in other industries, AI will be a force multiplier, with individuals being much more productive and using their knowledge in deeper and more insightful ways.
On the point of accuracy, I am not too concerned. We've been living in a 20-year search/Google world and the data provided is always filtered by humans. The list of information that is brought up must be scanned to identify the garbage from the tainted and then a horizontal review of them ultimately decides whether you trust any particular reference. Trusted sources, along with non-aligned consensus, are the way a good search is done today. AI recommendations will require a similarly defined trust process to determine if it has been 'hallucinating' or providing solidly researched data.
Reference: https://youtu.be/CWBTV0Qcxts
