​

In recent years, there has been a turn around in the structure through which healthcare is given and received. Advancement within the past ten to twenty years alone have allowed for a hyper-technicized and individualized form of healthcare. With technological modernization heading in the direction of artificial intelligence (AI), it appears that medicine and other health-related fields will change accordingly. While these alterations are going to face controversy, there is no doubt that the shift will take place, just like it did when phone calls or the Internet became easily accessible. There are three topics within this issue that must be addressed when considering how artificial intelligence fits into healthcare. The history of AI in medicine, the current and future uses of it, and the concerns related to its ethics are all relevant in determining the credibility of AI as a healthcare tool.

​

1. History

The earlier days of incorporating advanced technology into healthcare include some things that may be taken for granted today. In one of the first instances of AI in healthcare, Lawrence Weed created a technical program in 1965 named “PROMIS” that allowed for a computerized medical record (Altman, 1999). This was the beginning of the dependence on computers to use the same problem-solving methods that humans employ. Specifically for doctors, this tool proved useful as it provided them with a means to have previous symptoms and potential diagnoses right in front of them. It substituted handwritten methods that would require much more time and meticulousness. Following shortly after this, the 1970s became a time in which a string of programs were created that extended on the computer’s ability to create accurate diagnoses and predictions for each patient (Altman, 1999). This, and further endeavors to make computers more accurate in diagnosing, encouraged the image that computers were reliable and credible sources of information. A system like the ones from this time period created the beginnings of a framework that allowed for trust in technology.

​

II.   Present Day

There are numerous technologies that build upon the previous diagnosis related computers. While there are many different approaches to what AI can accomplish, it seems that much of the research is oriented towards severe diseases like cancer and heart-related illness (Jiang et al., 2017). One prime example is the IBM Watson for Oncology, which is a system that predicts treatments and can potentially aid in diagnosing cancer. An important point that the researchers make in their discussion is that the technology is not meant to replace any actual human relationships, namely the one between the doctor and the patient (Somashekhar et al., 2017). These forms of AI, while in their initial stages of life, are providing an alternative to medicine. They are producing the first opportunities for the removal of the “middle man”, or in this case, the doctors, nurses, and other healthcare providers that deal with patients every day. Present day incorporation of AI into patient care is a slow-moving process due to the ingrained concept that human relationships are key to quality care. In other words, despite effective and reliable advancements, the increased use of technology and intelligent computers is typically accompanied with care and assistance directly from a human healthcare provider.

​

III.   Concerns with AI in Healthcare

The current uses and potential future uses of artificial intelligence within the healthcare and medicine are far-reaching. They are capable of producing insight and knowledge that human healthcare providers may not be able to see. However, ethical concerns with the incorporation of AI into medicine have slowed down advancement. One commonly discussed issue is about which system of ethics should be applied so that the least damage happens. The three modes of ethics which are often discussed in AI politics are utilitarianism, deontology, and virtue ethics (Goldsmith & Burton, 2017). Much of AI runs off of utilitarianism, but it can’t be definitely said that this is the right way to make decisions. In utilitarianism, the pros and cons of every decision and outcome are weighed. The outcome with the least negative outcome is chosen to make a decision. For medicine and healthcare AI, treatments and suggestions about how to go forward with a patient’s conditions could be greatly affected depending on which of the three systems is used. It is assumed that the more the three systems are examined and understood, the easier it would be to use technology in practical settings where it is applied (Goldsmith & Burton, 2017).

The specific concerns with AI in healthcare stem from two issues. One is how well AI will be able to carry out the doctors’ jobs. Some say it can be done because doctors use a method called the pattern-recognition algorithm (Diprose & Buist, 2016). In this pattern the doctor acquires information about the patient, including history and symptoms, which are then compared to the knowledge the doctor has about diseases and known conditions. When the doctor finds a condition that matches, they will treat the patient based on what they think the condition is. IBM Watson does exactly this, and therefore the concern that AI is not accurate will likely be irrelevant within the next couple years.

​

The other concern is that AI will be able to replace doctors completely. AI is unique and distinct from doctors in that they do not sleep or eat, they do not take vacations, they can scan many medical reports at once to find diagnostic information, and they are overall more efficient. There is the likelihood that, other than a few trained clinicians, there will not be a need for doctors in the future because they are costly and less efficient than AI will be (Diprose & Buist, 2016). This could change the job outlook for many doctors because specialization will no longer be a valuable skill. AI in medicine is therefore a concerning topic for anyone who feels strongly about the value of doctors in healthcare.

​