A Fierce Healthcare article published today points out that, despite the excitement in the healthcare arena about Artificial Intelligence (AI), “there are few practical applications currently I use”. Although the desire to take advantage of the promise of AI is strong, what is the barrier and why isn’t it working better – yet?
To understand this, the healthcare field can simply look to its own past. If I said medications would revolutionize healthcare, you would certainly agree that medications would be a powerful tool, among others. But, you would also immediately recognize that “medication” is a very broad concept. “Medication” is an approach – delivery of treatment molecules to an affected site. We know all too well the length of time and cost it takes to refine what we mean by “medication” for a specific disease, and the specific medication for a particular disease usually will be distinct for that disease.
AI is similar. AI, too, is an approach – in this case, one where computers carry out tasks that require human intelligence. For each specific problem, considerable development can be required to adapt the right algorithms. Substantial AI capabilities already exist and new approaches are rapidly evolving. The limitation may just be our patience. You don’t expect one miracle pill to cure all diseases. Take the time to define each problem, build the AI solution needed to solve that problem, and – in time – AI will indeed pervade all of healthcare information technology.