With the development of artificial intelligence (AI) capabilities over the past years, many industries have integrated AI-based tools into their workflow to improve productivity. Healthcare is no exception; AI use in medicine ranges from managing patient data to improving the efficiency of diagnosis. In diagnostics, AI offers an opportunity to improve accuracy and patient outcomes while reducing delays. In this paper, I examine the viability of AI diagnostic technology via an analysis of existing research, including best-diagnosed conditions and common limitations. I then explore interview physicians regarding AI use and their work environment to identify potential gaps between the capabilities of AI and the potential for implementation given attitudes within healthcare. Overall, there is a slight positive correlation between the size of the sample set used to train the algorithms and their performance outcomes. There are also notable differences in both training size and performance based on the body system addressed by a model, which is indicative of shortfalls in research and diagnostic capabilities in certain areas of medicine. These findings signal a need to encourage the construction and use of more comprehensive datasets. This result is supported by the physician accounts, which demonstrated a general interest amongst clinicians in using AI in the future as well as concerns about its abilities and consequences on the healthcare industry in the status quo.