Skip to content

The future of ethics and biomedicine: An interview

In this radio show, Stanford bioethicist David Magnus and host Russ Altman discuss the ethical implications of using AI in health care.

AI may allow for the introduction of new tools and processes that could greatly benefit patient health. But its use also introduces many ethical concerns.

In an episode of "The Future of Everything" host Russ Altman, MD, PhD, spoke with Stanford biomedical ethicist David Magnus, PhD, about the impact of AI on the future and ethics of health care.

According to Magnus, there are four basic ethical issues that arise when introducing AI to affect care for patients. The first is hidden biases in machine algorithms that can influence the AI's clinical advice. For example, an algorithm that guides decisions on which patients go to the intensive care unit may be motivated by individuals' insurance status, thus placing the hospital's finances above patient care.

The second issue is the development of AI systems that are designed to perform in unethical ways. AI is only as unbiased as the data it is learning from. Therefore, if the machine algorithms are based off of data that suggests that developmentally delayed children respond poorly to organ transplants, for example, those biases will be continuously reinforced and validated within the medical care system.

Another concern is a lack of physician knowledge regarding how to interpret AI outcomes. Unfortunately, many physicians are unaware of the algorithms' limitations, causing them to use AI in ways that are inappropriate or problematic.

Finally, Magnus said, the addition of a third "party" into the doctor-patient relationship complicates things. Much of medicine's fundamental ethics are based on the assumption that there are two people: The physician and the patient. With with the introduction of AI, the physician's autonomy may be diminished, begging the question, who is really responsible for medical decisions?

However, according to Magnus, there are still many positives with AI. For example, in the field of genetics, there tend to be far too many data points for physicians to manage. "The potential for using algorithms to really help make decisions under a lot of uncertainty where clinicians don't have the expertise to do this is really going to be important," he said.

To hear more about Magnus's research on implementing machine learning in health care, have a listen.

Image by geralt

Popular posts