The Stanford One Hundred Year Study on Artifical Intelligence 2016 Report is a comprehensive report that accurately portrays the impact of artificial intelligence is already having on many modern facets as well as its potential future impact. Though seemingly correct in many of its arguments, one key area wherein the report over characterizes the impact of artificial intelligence is in the field of medicine, as it remains highly unlikely that artificial intelligence systems will be regularly be used by doctors despite the promise they show due to their inherent risks. One key area wherein the article states that AI can impact medicine is in the area of medical diagnosis. By training AI to learn the symptoms of various diseases through prior cases of diseases, the article postulates that soon it may be possible to use AI as a supplementary tool thereby allowing doctors to more accurately determine accurate diagnoses. Though this proposition seems logical, it presents two major logistical issues: the complexity of modern diseases and the accuracy of AI diagnoses. Unlike most fields, the field of medicine is multi-faceted. Diseases are not simply a factor of type and cause but are also affect people of different ages, sexes, nationalities and life-style choices and a vastly different manner. Currently, when a doctor is making a diagnosis regarding all of these issues, all of these seemingly complex factors must be appropriately taken into consideration when giving a diagnosis. Although with enough training data, it may be possible to create an AI system that is able to perform such a task, the ability to integrate so many different parts of an individual patients life when giving a diagnosis may prove for the introduction of AI into this specific field to be a bit difficult. In addition to this singular problem, most medical diagnosis are dynamic meaning that depending on how the patient responds to different types of treatment, the original diagnosis may or may not change. Although AI may be able to possible assist in the initial diagnosis of certain diseases, doctors through their rigorous medical training and their experience working in the medical field remain the best and most qualified decisions wherein a human’s life may possibly hang in the balance.
Another area where the report stated that AI could possibly find use in medicine was in the field of image processing. In the report, the authors state how AI could be trained to recognize tumors by using a training set of data. By then running this training set against actual data of images of various tumors, the AI would then be able to provide accurate analysis about whether a tumor proved to be dangerous to an individual. Although this technology is already seeing very limited usage, it is highly unlikely that the technology will become widespread due to the inaccuracies that may exist within the AI. Regardless of how much an AI system is trained on training data, it will never be able to perfectly predict with a 100% accuracy whether an individual’s tumor is harmful. In addition, such a technology would have a hard time keeping up with the rapidly changing landscape of modern medicine and would not always have to access of training data that it may need in order to ensure that it can provide an accurate analysis of the image. For this singular reason, it is highly unlikely that we see a widespread usage of AI within this specific field.

Although many of the points made in the article are valid, those on medicne cannot be. Due to the high risk nature of working with human patients, and the uncertanity that could possibly exist, it may take a long time for such technology to exist. In addition, as two cases will not be the exact same, there is unlikely ever to be a singular program that can be run to definitivily diagnose better than human doctors can as many diseases have enormous complexities.