THE RISE OF MEDICAL AI
Over the past few years we have seen a rise in AI programs that have an uncanny ability to make a medical diagnosis with more accuracy than humans. Increasing amounts of health data from apps, personal monitoring devices, and medical records have provided machines with as much information as possible rendering them more powerful than ever before in the field of medicine. Algorithms can also perform complicated tasks in minutes that take human doctors hours or even days to perform.
There is absolutely no doubt that AI will soon augment the efforts of doctors but will they ever completely replace them? That is the question we will tackle in this post by looking at typical arguments for and against the idea of AI replacing medical personnel.
THE DIMINISHING ROLE OF THE HUMAN DOCTOR
Even before AI took off one can argue that after the internet became commercially available the role of the human doctor began diminishing. Before the days of the internet the only real source of medical information was a doctor. Nowadays there are entire groups of people who can do their research and keep themselves healthy enough to where they hardly need a doctor at all. These groups of people will only grow larger.
Also, back when the medical industry was still young, doctors were able to play a major role in patient care by providing an ability to think laterally and analyze each patients individual case and provide empathy for the patient that many people believed was crucial. Over time though, medical practices became more crowded and more focused on making a profit and less focused on diagnosing and treating the problem at hand.
If you have been to the general practitioner lately, I’m sure you’ve noticed that they only have a few minutes to dedicate to each patient before quickly writing a prescription if necessary and telling you to come back if there are any more problems while billing your insurance company each time you return. These doctors are also compensated by the drug manufacturers for each prescription they write. Combine that clear conflict of interest with the fact that doctors no longer have the time to give individuals the attention they need and the fact that medical science is moving forward so fast that they don’t have time to keep up with the research and it is no wonder we see their role gradually decreasing with each passing year. The point about empathy is also a faulty one which we will address in the next section.
IS EMPATHY REALLY NECESSARY?
Proponents of human doctors never relinquishing complete control to AI cite the human’s ability to empathize with their patients as a primary reason why machines can’t do the job. However, after the multi variant analysis is performed this statement simply doesn’t hold up. While empathy is crucial when dealing with infants or disabled, people most patients value diagnostic accuracy and a solution to their problem over an emotional connection with their physician.
A recent Harvard review article challenges the notion that because current machines cannot show empathy they cannot replace doctors. The authors of this article also argue that compassion is not necessary in other lines of professional work. They even cite certain human attributes that machines have a hard time with like creativity and judgment and claim they are also not necessary in medicine. Regardless of the fact that creativity and empathy are not really needed to become a doctor there is no doubt AI will become more creative and compassionate with time. The idea that AI will be comprised of emotionless programs and can never attain more human like qualities is inherently flawed.
CLAIMS THAT AI WILL MERELY AUGMENT DOCTORS INSTEAD OF REPLACING THEM
A number of people with vested interests in the medical profession have claimed that AI programs will only be able to augment the efforts of current personnel and will never be able to fully replace them. Most of them argue points that we have already begun to pick apart in this post but there is a little more to unpack.
DeepMind researcher Dr Alan Karthikesalingam recently appeared at SIngularity Universitiy’s Exponential Medicine conference in San Diego in early November. There he presented an idea that the future of medicine is a kind of physician-patient-AI golden triangle in which both parties benefit from the introduction of AI without being completely replaced. DeepMind is the company that created Alpha Zero, the powerful neural network that mastered chess in only a few hours. That particular AI has mostly moved on to folding proteins and DeepMind continues to pursue ambitious goals in the field of medicine.
Opponents of completely autonomous AI doctors point out that it has been the role of healthcare professionals to be the main gatekeepers to ensure that new technology and new treatments are safe and effective before they are introduced. However, as stated earlier the role of the doctor is diminishing and there are serious conflict of interest problems that cannot be denied.
Highlights of the recent Exponential Medicine Conference.
THE ROLE DEEPMIND IS PLAYING
Companies like DeepMind are working to create programs that supplement the efforts of physicians. Even if these supplemental programs ensure that doctors never make an incorrect diagnosis they will prove unable to get rid of human bias and conflicts of interest. When this line of thought is pondered for a while it becomes clear that whether we like it or not AI will replace doctors at some point in the future.
Dr Karthikesalingam stressed at the conference that the goal isn’t to replace medical practitioners, it is to optimize their performance while taking care of the more menial tasks for them. This is a good forecast for the near future but not for the far future. In the far future when AI reaches a general intelligence threshold there will be no stopping it from pushing humans out of the industry. Overall this will be a good thing provided certain regulations and safety measures are put in place. Furthermore, to be fair to DeepMind, they were bought out by Google not too long ago which means despite what they say or claim there is no good way to truly know what their actual long term intentions are. Google doesn’t exactly have a shining reputation when it comes to exploiting people. Only time will tell.