Will Artificial Intelligence Replace Doctors?

Image from Newsweek

The era of human doctors is coming to an end.

In the near future, it may be a machine, and not a human being that treats you in the hospital. At least, this is the reality that AI, artificial intelligence, is making come true.

There is a major problem that exists with having humans be the medical providers: human…error.

No matter how accurate humans are, it is part of being human to make mistakes, which is mind boggling when you consider how powerful the human brain is.

A 2016 study from Johns Hopkins University found that medical errors may be the third leading cause of death (this is a debatable issue with many disputing their claims, but a discussion we must have nonetheless), right behind cancer and heart disease, with an estimated quarter million patients dying annually from medical error.

Who can save us from this terror? Introducing medical machines.

Dr.StickFigure Youtube Channel

You may think there’s no way machines could replace a complicated job like a physician, but this is something that is already happening. There are more monthly visits to WebMD than there are to actual doctors. Smartphone Apps like Ada put the power of diagnosis in your pocket, using machine learning to figure out what disease a person has based on symptoms. In China, the first AI computer has passed medical board examinations and the first AI- assisted treatment center has already opened where doctors say computers both help them, and compete for their jobs.

Most welcome machines replacing tedious jobs like those in transportation, delivery, and manufacturing, but you’re probably thinking there’s no way a computer could have the empathy, creativity and judgment to replace a doctor. This could not be further from the truth, however. Researchers Richard and Daniel Susskind disprove this notion in their book, The Future of the Professions, where they argue that if tasks are broken down into component parts, the creativity and empathy are no longer needed.

For example, researchers at Stanford found that a machine learning program can diagnosis skin cancer just as accurately as a panel of dermatologists, and another machine learning algorithm can predict who’s at risk for heart disease much better than doctors can. There’s even an algorithm that can look at your Instagram account and predict with 70% accuracy whether or not you were diagnosed with depression in the past three years. Now that’s… scary.

How does a machine do this? Basically, the machine is fed thousands of pictures of what is known to be a cancer and what is not. The machine examines these pictures and picks up on certain trends of what disease looks like. And over time, with failure or success, the machine continues to learn and get better. For example, if IBM Watson makes a medical mistake in New York, it can teach itself not to repeat the same mistake globally across all networks it functions. If a doctor in new York makes a mistake, it will be impossible to pass that information on to all doctors across the planet.

Using machine doctors is also much cheaper. With the upcoming doctor shortage and a collapsing healthcare financial system, many could turn to the machines for medical care, whether or not it’s a better option.

Training a doctor takes thousands of dollars and years of hard work…trust me I know. But a computer can learn and diagnose much faster. A machine can calculate all the drug interactions and possible treatments much faster than a mere human doctor can. With medical knowledge doubling every three years, it may be hard for humans to keep up.

Also, a computer doesn’t have eyes that get tired after looking at thousands of x-rays, it doesn’t need to sleep at night, and doesn’t ask for a vacation…which would be so nice right now (doctors and medical students across the world can relate).

Technology companies have noticed the power of machine learning and are hoping to take advantage of it. The Chan Zuckerberg Initiative has invested in an AI company, Meta that hopes to analyze all the research studies being done and find trends that it can teach to doctors, since its impossible for anyone to read them all. IBM Watson has become quite famous and Microsoft is trying to use machine learning to find better cures for cancer.

But should we let machines take over? Do you want a cold, heartless machine telling you that you have 6 months to live? One of the most important parts of the field of medicine is the human aspect, the connection that happens between the physician and patient as a result of the care, the trust, and the servitude.

The ideal scenario has machines and doctors working together for the betterment of the patient. For example, IBM Watson is already helping doctors find treatments for cancer than they might have missed

A radiologist can use Watson to make a better reading of a CT-scan, which ends up with fewer misses and errors. I still think doctors, and the human connection they form with patients will remain a critical part of the healthcare system. That being said, however, doctors will have to adapt to be focused more on the human aspect of medicine, which in my book, is the most important part of being a doctor anyways. The merger of doctors and machines has already begun.

What do you think? Do you want a robot replacing your doctor or does that future sound horrifying?