Preface

Human history is full of instances where new inventions have created a sudden, significant, and lasting disruption. In a somewhat gradual and even stealthy way, artificial intelligence (AI) and machine learning (ML) are becoming part of our everyday lives, changing things in both predictable and unpredictable ways. This "randomly systematic" adoption process is putting humanity face to face with something never previously directly known to our civilization: an intelligence that may (and likely will) exceed our own.

It is fair to say that most people are not fully aware of current (and thus future) benefits, limitations, and threats related to AI/ML adoption. Within healthcare and medicine in general, there is little awareness of what AI/ML actually entails and what it is capable of at this time. It is this current state that will serve as our "starting point" in the emerging debate on AI/ML in medicine, including its integration, projected influence, and a variety of other considerations that are not all that different from other past technology adoption paradigms.

Like all other transformational human inventions, the emergence of AI/ML is a culmination of various simultaneous developments, often parallel and co-dependent, but also unpredictably synergistic, that ended up amalgamating to facilitate computational processes that approximate the "functional outcomes" of various human logical processes. Among the advances that were required for AI/ML to enter the mainstream were modern integrated circuits, higher computer processing speeds, ability to deploy parallel-processing capabilities, greater amounts (and lower cost) of computer memory, software engineering knowledge, and the ability to harness the power of the Internet to gather vast amounts of high-density data in an efficient and highly structured manner.

This new "artificial intelligence" phase in human history represents a confluence of multiple factors uniquely coming together to change our civilization forever. Although significant threats and opportunities exist in relation to real-life implementations of AI/ML in health care and beyond, a tremendous amount of promise and positive developments may also be realized. We are at a crossroads, and the outcome of any decisions made "right now" will heavily depend on whether we (i.e., humanity) make the right collective decision, at the right time, and for the right reasons. Although the gravity of this historic moment may not have yet become apparent, we will have to live with its consequences.

When implemented optimally, AI/ML has the potential to result in vast improvements in healthcare efficiency, workflows and other related processes, patient safety, and overall clinical outcomes. Specific benefits of AI/ML in the clinical realm include better, more accurate, and faster diagnostics; early disease detection, especially as it applies to cancer, cardiovascular, and genetic conditions; dynamically updated clinical guidelines, informed by actual patient outcomes and supplemented by real-time

outcome data; and many other as yet undefined enhancements and advances. Specific threats related to AI/ML include workforce displacements (due to redundancies created by AI-based efficiencies); loss of human autonomy (due to "outsourcing of decision-making" to AI-based systems); and the propagation of various deleterious systemic biases (due to AI system reliance on potentially biased data feeding its ML algorithms).

As the early, more rudimentary capabilities of AI/ML continue to grow and mature, so will the diversity of the associated clinical applications. With further enhancements in hardware, software, and implementation infrastructure, increasingly complex areas (and problems) will become amenable to AI's general "scope of abilities" and influence. This will gradually expand into highly sophisticated systems and areas, such as social sciences and health care. In this collection of chapters, we discuss current trends and future developments related to AI and ML across medical and surgical specialties.

> **Stanislaw P. Stawicki, MD, MBA, FACS, FAIM** Department of Research and Innovation, St. Luke's University Health Network, Bethlehem, PA, USA

Section 1 Introduction
