INTEGRATING A.I. INTO MEDICAL PRACTICE Part 2

OPPORTUNITIES AND PERILS

A decade after my retirement, scientists with the advancement in computer science and microchip technology were able to invent such a reasoning and super intelligent machine tool called artificial intelligence (AI). These tools are still in the developmental stage, would be able to do anything with their limitless memory capacity, lightning speed and far superior IQ level and knowledge than humans.

I thought how lucky the junior doctors are, they would no longer have to practice medicine with uncertainty and with the fear of malpractice.

I have read everything I can put my hand on about AI. As I understand it, AI reasoning tools are built and algorithmically programmed machines or platforms with human intelligence.

Two types of AI have been pursued for medicine: machine learning, which I don’t quite understand, and natural processing, which is the ability of advanced computers to transform human language and texts into machine readable data that reflect the intent of the language. However, the more I read about AI the more concerned and fearful I became for the following reasons:

It appears that AI based reasoning tools are not simple and intelligent reasoning tools that can be used by doctors to support and verify their medical decisions, thereby eliminating medical uncertainty in practice of medicine.

AI machines could have their own mind, they could hallucinate just like a mental patient, see non- existing facts, and present like facts beyond the control and understanding of scientists and programmers.

AI powered reasoning tools do not have a soul, emotions and feeling of empathy. They just give medical knowledge and advice with cold and dry machine language.

If AI machines completely take over the cognitive works of doctors and eliminate human emotions and human touch in medical care, would patients be satisfied? Would doctors remained committed to lifelong learning, and moral and ethical principles of practice of medicine?

Can ethical decisions about patients be made by standard medical algorithms since it could vary from patient to patient?

Could medical privacy be assured by AI-powered machine tools?

For all the above reasons, I believe medical societies and more importantly medical educators must get involved and lead the way integrating AI into medicine and medical practice together with AI scientists.