Dr. Scott Silverstein on Technology, Human Judgment, and the Future

Artificial intelligence has become one of the most discussed technologies in the world. It is transforming how we work, how we process information, and how we think about the future of medicine.

But should we be afraid of it?

That was one of the questions explored in a recent interview with Dr. Scott Silverstein, a pioneer in medical informatics with more than six decades of experience working with computers, computer science, and the evolution of technology.

Having witnessed the transition from early computers to electronic medical records and modern artificial intelligence, Dr. Silverstein brings a perspective that is both informed by history and grounded in practical experience.

His message is not that AI is harmless or that every technological advance will be beneficial. Rather, it is that we should approach these tools with appropriate safeguards, respect for their limitations, and confidence in the importance of human judgment.

From a High School Computer to Medical Informatics

Dr. Silverstein’s interest in computing began in high school, when a mathematics teacher purchased a Digital Equipment Corporation PDP-8S computer in 1969 for student use.

At a time when personal computers were still years away, access to a computer was an unusual opportunity. Dr. Silverstein became deeply involved in computing while also pursuing his interest in medicine.

During medical school at Boston University, he worked on computer-related projects, including interfacing a nuclear medicine camera with a PDP-11. He later completed an internship and residency before discovering that the National Institutes of Health was funding postdoctoral fellowships in medical informatics.

He was accepted into Yale’s postdoctoral fellowship program, where he spent two years working on computer-related projects around Yale New Haven Hospital, including projects involving invasive cardiology and cardiac surgery.

He subsequently joined the faculty at Yale School of Medicine.

His career illustrates something important about technological progress: the tools may change dramatically, but understanding how to use them effectively requires knowledge, curiosity, and experience.

The Syndrome of Inappropriate Overconfidence in Computing

During the interview, Dr. Silverstein described a phrase he coined while electronic medical records were becoming more prominent:

The syndrome of inappropriate overconfidence in computing.

He developed the phrase after hearing repeated predictions that electronic health records and decision-support systems would revolutionize medicine.

The problem, he explained, was not that these systems were useless. They could make certain tasks easier and help physicians manage information. But the predictions surrounding them were often far more confident than the technology justified.

Electronic medical records have made important contributions to healthcare. At the same time, they have introduced challenges, including complicated interfaces, workflow problems, and situations in which the technology itself can contribute to error.

The lesson is not to reject technology. It is to recognize that a useful tool is not necessarily a revolutionary solution to every problem.

That distinction becomes especially important when discussing artificial intelligence.

What Is Missing From AI?

One of the most interesting points in the interview was Dr. Silverstein’s explanation of what artificial intelligence still lacks.

Human experience and judgment.

AI systems can process enormous amounts of information, identify patterns, and assist with complex tasks. But Dr. Silverstein explained that they are attempting to emulate the human mind, and he does not believe that a true emulation of the human mind is likely to happen in his lifetime.

There is a difference between simulating aspects of human thinking and possessing the full range of human experience.

In medicine, that difference matters.

A complex medical problem may involve not only test results and imaging, but also a patient’s emotional state, social circumstances, personal history, and the judgment developed through years of clinical experience.

These are not always matters that can be reduced to a set of data points.

Dr. Silverstein believes AI can be a valuable tool for correlating large amounts of information. He uses it himself for that purpose. But he also makes an important distinction: AI can assist with analysis without necessarily making the final judgment.

AI as a Tool, Not a Replacement for Human Judgment

During the interview, Dr. Silverstein described using AI to help compare CT scans and ultrasounds.

The technology provided a precise analysis and saved him the time required to manually lay out and compare the studies. However, he emphasized that the tool did not make the medical judgment for him.

That distinction is central to the responsible use of AI in healthcare.

A system that helps organize information can be incredibly useful. A system that is treated as infallible can create new risks.

The goal should be to use AI to support human expertise, not to assume that the technology eliminates the need for it.

Every Technology Has Benefits and Risks

Dr. Silverstein also discussed the broader field of social informatics, which studies the social impacts of new information and communication technologies.

He explained that research in this field has identified a recurring pattern: new technologies can produce useful and beneficial effects while also creating unexpected negative consequences.

Electronic medical records, mobile phones, and artificial intelligence all illustrate this principle.

Mobile phones have brought enormous benefits, even though they can also contribute to distracted driving. Electronic medical records can improve access to information, even though poorly designed systems can create difficulties and errors.

The existence of risks does not mean that a technology should be abandoned.

It means that we need to understand the risks, improve the technology, and establish safeguards that make its benefits more accessible and its harms less likely.

Should We Be Afraid of AI?

The interview began with a discussion of recent warnings that artificial intelligence could pose an existential threat to humanity.

Dr. Silverstein’s response was measured. He acknowledged that any tool can be used for harmful purposes, and he recognized that technology has sometimes produced adverse effects.

But he did not believe that AI should be approached as an inevitable doomsday scenario.

Instead, he expressed optimism about the ability of people who understand information and information science to use AI to manage and draw inferences from large information sets.

That is a more practical way to think about the future.

Fear alone does not improve technology. Understanding, research, safeguards, and responsible decision-making do.

The Human Future of Technology

Perhaps the most important message from Dr. Silverstein’s interview is that technological progress does not eliminate the value of human experience.

The tools available to physicians have changed enormously since he first began working with computers in high school. Yet the need for judgment, responsibility, and an understanding of the people behind the information remains.

AI may help us process information faster. It may help us identify patterns that would otherwise be difficult to see. It may make certain tasks more efficient.

But those benefits do not mean that human beings have become unnecessary.

They mean that we have new tools to learn how to use.

As Dr. Silverstein put it, we should not be afraid of AI. We should control it, capture it, and make it as safe as possible.

That is not a promise that every technological development will be perfect. It is a reminder that the future is something we participate in shaping.

And when it comes to medicine, the best future is not one in which technology replaces human judgment.

It is one in which technology helps human judgment become more informed, more efficient, and more capable of serving patients.

Interested in my book? You can buy it on Amazon!

(𝘛𝘩𝘦 𝘷𝘪𝘦𝘸𝘴 𝘦𝘹𝘱𝘳𝘦𝘴𝘴𝘦𝘥 𝘪𝘯 𝘮𝘺 𝘱𝘰𝘴𝘵𝘴 𝘢𝘳𝘦 𝘮𝘺 𝘰𝘸𝘯 𝘢𝘯𝘥 𝘥𝘰 𝘯𝘰𝘵 𝘳𝘦𝘱𝘳𝘦𝘴𝘦𝘯𝘵 𝘵𝘩𝘦 𝘷𝘪𝘦𝘸𝘴 𝘰𝘧 𝘮𝘺 𝘦𝘮𝘱𝘭𝘰𝘺𝘦𝘳 𝘰𝘳 𝘢𝘯𝘺 𝘰𝘳𝘨𝘢𝘯𝘪𝘻𝘢𝘵𝘪𝘰𝘯.)