After the 2016 incident in which an artificial intelligence defeated a Go champion, there has been a lot of discussion in the medical community about whether advances in artificial intelligence will replace doctors as a profession. While diagnostic systems such as IBM’s Watson are rapidly advancing, the role of the doctor is still important, and AI will serve as an auxiliary tool.
In 2016, the world was once again in awe of AI’s capabilities. In Go, known as the world’s most difficult board game, an AI defeated the world’s number one Go player, Lee Sedol. AI had already lost to chess and trivia, but Go was still considered the last bastion of humanity. However, this result shocked many people. For a while, the world was filled with stories about AI. News stories about how society will change in the future due to the development of A.I., the importance of coding education in education, and the Fourth Industrial Revolution became a regular topic in politics. In the aftermath of this, the medical community has also been talking about the future of medicine with the development of A.I. for some time. In this regard, Watson was developed by IBM in the United States. Watson’s ability to make accurate and fast judgments and penetrate deep into the medical field has led to many bleak predictions about the future of the profession, both inside and outside of the medical field. But will the future of medicine be one where doctors will disappear as AI advances, and if not, what will the future doctor look like?
Before we get into that, let’s take a look at what AI looks like in healthcare today and how far it has come. Let’s take the case of Watson, the current hot topic in medicine, as an example of where AI is today. Watson has already been implemented and used in a few hospitals. Watson is a case-based artificial intelligence diagnostic developed by IBM that has yet to catch up to humans in terms of accuracy, but it’s blazing fast in terms of time to diagnosis. This is because Watson can read more than 10,000 papers in a matter of minutes, whereas a typical doctor can read about five papers a day. It accumulates the knowledge it gains from these papers and also accumulates a large number of patient cases, which it then correlates with the patient’s symptoms to diagnose the disease, the same way doctors do. In fact, some hospitals have been using Watson since 2016. Although Watson is a Western technology that is not as accurate as its Asian counterparts, there have been cases where Watson has saved patients’ lives. However, Watson is not yet at a point where it can be relied upon to diagnose diseases, and it is still in the research phase. Right now, it only checks to see if the conclusions made by doctors and Watson agree. Even for the top five cancers, Watson is less than 90% accurate. This kind of deep learning technology is new, and it’s still a few years away from being used in medicine, as it’s only just begun to beat board games like Go.
Given how far we’ve come in the field of medicine, what is it about AI that makes the future of medicine so bleak? We can look to Yuval Harari’s ‘ Homo Deus ‘ for evidence. The author focuses on the role of doctors, arguing that their main job is to properly diagnose illnesses and suggest the best treatment, which will eventually be replaced by AI. In fact, AI can read a lot more papers than doctors, and it can make diagnoses quickly and accurately without making mistakes. Furthermore, given the huge amount of time and money it takes to train a single doctor, the role of doctors is predicted to shrink even further. So, even if doctors aren’t completely replaced by AI, the number of doctors is expected to decrease dramatically. In fact, this concern has been raised much earlier. If we divide doctors into two categories, surgery and internal medicine, the former is already showing better results than human doctors thanks to advances in surgical equipment such as DaVinci Surgical. In the case of internal medicine, further advances in diagnostic programs such as Watson may mean that doctors will no longer be needed.
So, will doctors really be replaced by AI? I don’t think doctors will ever be replaced by AI, and there are two reasons why.
The first is that AI can’t make a perfect diagnosis for every disease. Of course, it’s true that technologies like Watson are now approaching 90% accuracy for the top five cancers. But there are dozens of different types of cancer, and there are many other diseases. AlphaGo may have beaten Lee Sedol at Go, but AlphaGo is a program built for Go. What would happen if you gave AlphaGo a new board game? Of course, like Go, it could be trained thousands and tens of thousands of times to build an algorithm, but until then, it would lose to humans. For AlphaGo, a new disease is like a new board game. It would need tons of cases to learn the “rules,” and since it’s new, there’s no data. Of course, doctors are also vulnerable to new diseases. But that’s why doctors do research on new diseases, so they can develop antibiotics or vaccines, or understand the cause of the disease so they can take preventive measures. Humans can do more than AI. Of course, for common diseases, AI will one day be able to make clearer and faster diagnoses than humans. But we shouldn’t worry about this at all. AI was developed to help humans in the first place, so doctors can use it to their advantage when diagnosing diseases. When the robotic surgical device DaVinci Surgical was introduced, many surgeons were worried that they would be replaced by robots. But in the end, only a surgeon can handle it, and surgeons now use it as an “assistant” when performing delicate surgeries. The same is true for Watson. In the future, Watson may be able to diagnose illnesses and suggest appropriate treatments better than humans, but it’s still doctors who input patient data and write the papers Watson references. Decades from now, Watson will still sit on a doctor’s desk, acting as an assistant to help them diagnose patients.
The second reason is that we need to rethink the role of the doctor. When you think of a doctor outside of the medical community, you probably think of a doctor in a white coat examining a patient, or a doctor in scrubs operating on a patient. But that’s only one side of the picture. When patients are sick, they trust their doctors the most. That’s why doctors need to be able to treat patients with extra care and inspire trust when they make rounds. This belief is deeply embedded in people’s minds. But does putting a white coat on a machine make it trustworthy? This is a controversial topic, but even after evolution and the big bang theory, many people still believe in creationism and still go to church every Sunday. Similarly, the profession of a doctor will remain a symbolic one, even if AI takes over the role of diagnosis. Furthermore, the place of doctors in society is also important: they play an important role in determining health policy and coordinating health insurance. The role of doctors is not just limited to hospitals. This is where AI cannot take over. Doctors also share new treatments and research through conferences. Sure, Watsons around the world could read all the papers as long as they have an internet connection, but it’s still doctors who write these papers, and it’s doctors who decide the direction of future research, so this academic exchange will continue.
Of course, there are counterarguments. In the distant future, if A.I. becomes more advanced and more creative than humans, and is able to study diseases and write papers, there will be no need for doctors at all. It is also argued that if A.I. is able to replace the ability to communicate with patients, doctors may eventually be replaced in all aspects. For example, in the 2014 movie Her, we see an A.I. with human emotions, and we see how people have a deeper relationship with it than they do with humans. It has been suggested that in the future, AI will be able to communicate with patients better than human doctors and that doctors may no longer be needed.
However, this discussion paints an overly pessimistic picture of the future. Even during the Industrial Revolution, there were fears that machines would replace humans, but as automation progressed, people found new jobs and a new field of service industries emerged. The same is true for AI. At the current rate of development, it’s unlikely that AI will take over the hospitality industry in the next few years, but we shouldn’t worry too much if it does. Just as people in the past couldn’t have predicted the various jobs of the 21st century, we can’t predict exactly what the jobs of the future will be. The same goes for doctors. Of course, the doctor of tomorrow will look different from the doctor of today. But this can be seen by comparing doctors of two hundred years ago to those of today. Just two hundred years ago, doctors were more like shamans than healers. There was no concept of hygiene, and people believed that if you were sick, drawing blood would make you better. Also, just 30 years ago, there was no concept of clinical trials as we know it today. Case reports were all that existed, and doctors with more experience had an advantage. Could they have imagined the doctors of today? Similarly, the current advances in AI don’t mean that we won’t need people to cure diseases. As long as there are people who are sick, there will always be people who can fix them. While it’s impossible to predict exactly how doctors will coexist with AI in the future, it’s a stretch to think that the profession will be eliminated by AI. Even if AI has better communication skills than humans, doctors will have another role to play.
The reason why Yuval Harari’s ‘ Homo Deus ‘ discusses the specificity of doctors and the possibility of their replacement by AI is because many people are anxious about the development of AI and the loss of their jobs. Of course, there will be some jobs that will be completely eliminated by AI. However, the development of A.I. means the birth of a new class of jobs in the era of the Fourth Industrial Revolution. We need to be flexible and adaptable to the transitional period. In the end, I would like to say that the future doesn’t have to be so bleak because of AI. This is a personal opinion, of course, but at least the current debate around AI has been overheated, and the future we live in will not be this bleak. On the contrary, I believe that AI will enrich our lives and create a more stable society.
