Abstract

Artificial intelligence (AI) is being widely used, with trial systems already in use in the medical field, where there are high expectations for its potential for use in the future. Studies have shown that AI diagnosis is almost as accurate as that of human experts. However, problems with the use of AI in clinical medicine must be identified and carefully examined to insure that it can be used safely. Herein, I illustrate various concerns related to the use of AI and show how it can be put to practical use in clinical medicine. The concerns about the use of AI include the following: differences in expectations for AI, failure to correctly diagnose disease because of a lack of data available for learning, and over dependence by physicians using it to make a diagnosis. AI can also be useful to hospital staff other than physicians. We should not leave the business of medicine and healthcare to AI systems, but use it to assist with decision-making and to raise labor efficiency and effectiveness.