Machine whisperers

Do you know what your doctor will be specializing in soon? She might be learning how to interact with machines, not with people. It is predicted that in the future, the doctors will have to adjust to a healthcare environment that is dominated by artificial intelligence (AI).

According to the article Medical Education Must Move From the Information Age to the Age of Artificial Intelligence by Steven A. Wartman, MD, PhD, and C. Donald Combs, PhD, here are the changes that we will encounter soon:

1. Health care will be provided in many locations since data storage/processing will be accessible in real time everywhere.

2. Health care will be provided by teams of physicians, nurses, social workers, therapists, home health aides, support groups, families, etc.

3. Health care will be guided by large data sets and artificial intelligence.

4. Physicians will be cognitively and physically outperformed by devices.



To understand some of the first applications of AI in clinical practice, watch this Google talk:
 


There are many aspects of the AI - dominated medicine that can be discussed; however, I am thinking about three imminent consequences.

First, the medical education would no longer focus on basic sciences. Medical education will prepare the future doctors to work with the machines and AI, which will churn out analyses of gigantic data sets. Decisions on life and death will be taken based upon algorithms. The docs will become “machine whisperers”. 

Second, docs will mostly manage machines and teams to deliver various aspects of the patient care. The landscape of educators in medical schools will change from this of basic and clinical scientists/practitioners to this dominated by specialists in informatics and programmers.

Third, since the teaching of basic sciences knowledge will dwindle, the medical doctors might not be able to replicate the logical sequence of differential diagnosis. They will know how to work with the machines and AI, but they would be in the dark about the basic reasons behind the causes for pathological conditions and proposed treatments. After all, such information would be accessible immediately through technology, as the docs need it. 

This situation reminds me of a conversation I had more than 20 years ago in my research lab. As a PhD student, I loved talking to one of my senior researchers. She had more than 30 years of experience and had been one of the first laboratory technicians applying molecular biology approaches in cancer research. She knew how to isolate RNA and DNA, she prepared the perfect protein lysates. She knew why each step in a protocol was included, why particular chemicals were used, in what proportions, etc. At the time, we were discussing the new kits that made the life of all researchers much easier. The kits came with protocols that specified, “use 5 ml of solution A, mix for two minutes, then add 20 ml of solution B”. Most of us, the students, blindly followed the directions without any understanding of the guiding principles behind the procedures. We did not know what solutions A and B were, why we had to mix them and why timing was crucial...

Could the same intellectual and informational blindness affect our medical doctors? I believe, it can. They may become no more than “machine whisperers” and managers of teams. Laboratory/clinical research, individual creativity and logic might be abandoned little by little. The trend has already started. Consolidation of research allows only the biggest and financially strongest players to stay in the game (think of whom the big investors fund today). There is no longer room and opportunities for the smaller, less financed schools and researchers to continue exploring new ideas. The level of creativity and initiative will change, as less and less individuals are allowed to develop and practice their imagination and innovation skills.

For a fictional interpretation of how AI and humanity may interact check out this post.