Artificial Intelligence and the Christian Physician
To speak of artificial intelligence (AI) conjures dazzling images of an electronically reconfigured future managed, if not dominated, by calculating, thinking, autonomous machines. Realistically, AI has the potential to deliver numerous useful benefits to medical practice, especially as progress in medical science and healthcare delivery rely increasingly on digital technologies to store and analyze huge
by, William P. Cheshire, Jr., MD
To speak of artificial intelligence (AI) conjures dazzling images of an electronically reconfigured future managed, if not dominated, by calculating, thinking, autonomous machines. Realistically, AI has the potential to deliver numerous useful benefits to medical practice, especially as progress in medical science and healthcare delivery rely increasingly on digital technologies to store and analyze huge data sets. The health information in the human genome and the scientific content of medical journals, for example, exceed the capacity of the human brain to recall, interpret or keep up with exponential advances. AI promises to bridge that gap. Proponents are calling AI the fourth technological revolution following the neolithic transition to agriculture, the industrial revolution utilizing mechanized production and new sources of power, and the digital revolution based on computer processing of digital information.
How should followers of Christ assess the fast-approaching promises and potential perils of AI? In particular, how should healthcare professionals not only think about thinking technologies that empower medical practice, but use them in ways that are consistent with a Christian understanding of medical professionalism and that honor the Great Physician?
When envisioning AI, nothing has enticed our imaginations more than science fiction. Some of the most interesting tales portray artificial intelligences either as all-knowing servants ushering in a new era of enhanced prosperity or as dangerous threats impassively intent on enslaving or ruthlessly replacing humanity. These stories are highly entertaining and raise provocative questions about what it means to be human. The prophecies of science fiction, while intriguing, fall short of providing satisfying answers to the questions raised, as the vast majority take place on the stage of a fantasy worldview in which God does not exist. The ethical assessment of AI must place it within the framework of reality.
The Future Now
AI is already upon us. From search engines to facial recognition technology, AI has entered into our everyday lives, tracking our habits and preferences, detecting our emotions, completing our sentences, offering product recommendations and directing our choices. Soon it may be driving our vehicles. Considering how thoroughly the internet and smartphones have changed our way of life, the potential impact of AI on human affairs is profound. It is also incompletely predictable.
In medicine, among familiar AI applications is natural language processing, which generates clinical notes from speech. Machine learning algorithms mine vast quantities of patient data to detect patterns of disease. Artificial neural networks utilize deep learning to analyze patients’ radiographic, histologic or morphologic images, and they are gaining in diagnostic accuracy. AI has the potential to identify disease earlier, when it may be more easily treatable, predict an individual patient’s response to specific medications and accelerate vaccine discovery. These are but a few examples of recent developments that could greatly benefit patients.
The potential for AI to improve health is a praiseworthy goal in keeping with the purpose of medicine. Insofar as AI applications can assist physicians in their calling to benefit humanity by diagnosing and treating illness and relieving human suffering, AI is a welcome instrument. Related to this is the potential for AI to improve efficiencies in access to medical knowledge, medical diagnosis, identification of safe and effective treatment choices and the work of detailed clinical documentation. AI as an intermediary between the patient and the human healthcare professional may also improve access to care. In that role, the line between resource and caregiver begins to blur.
Assessing Moral Responsibility
AI is a type of technology, and technologies can be used either for good or bad purposes. Understood from this “dual use” perspective, AI is morally neutral. The consequences of the deployment of technologies, however, are not always evenly divided among good and bad outcomes. The way technologies are designed can constrain human choices, promote specific habits of use or, in some cases, introduce potential harms that are disproportionate in magnitude or duration in comparison to the beneficial effects. For a very powerful or potentially irreversible technology, all of these aspects must be considered in the ethical analysis guiding decisions about appropriate use or restraint.
The Christian understanding of human nature also encompasses a dual aspect. All human beings are created in the image of God and are designed by their Creator to have inestimable worth (Genesis 1:25-26). At the same time, all human beings are fallen. Human nature is both splendid and sinful. Scripture encourages us to seek righteousness (Matthew 6:33) while also reminding us that all have sinned and fallen short of the glory of God (Romans 3:23).
These dual aspects of technology and humanity converge at the point where humans use technology. Whether technology is applied to good or bad purposes depends on human motives, which may be virtuous or unrestrained, altruistic or selfish, undivided or mixed, responsible or reckless.
AI differs from previous technologies in the way its actions are further removed from human decisions. Machine intelligence is designed to function autonomously. Various degrees have been defined, depending on the engagement needed from the human operator to monitor performance. These conditions include whether the machine can operate in certain situations or in all circumstances and whether the human has the option of overriding the machine and asserting control. For AI, the chain of causation between the human designer or operator and the machine’s effect is indirect and may be opaque. For the most advanced AI technologies, it may be inaccessible and untraceable.
This means that for AI operations moral responsibility is ambiguous. When the AI diagnoses cancer, it may be unclear whether or to what degree the physician operating the device deserves credit. When the AI commits a medical error, it may be impossible to determine where blame is due. Was the missed cancer the fault of the physician, the AI manufacturer, the AI programmer or the AI itself? Can a machine apologize and mean it? A machine might be programmed to mimic the expression of remorse, but can it feel genuine remorse? Can it truly care about the patient?
Philosophers debate whether machines with sufficient processing speed could ever acquire consciousness and, with it, the ability to understand the moral significance of their actions and thus incur moral responsibility. Some AI enthusiasts hold to a transhumanist philosophy that looks forward to a hypothetical future when AI has advanced to a state in which it not only mimics human intelligence convincingly, but exceeds it, and might even replace it. The bioethicist Julian Savulescu speculates that, “Humans may become extinct.… We might have reason to save or create such vastly superior lives, rather than continue the human line.”
The Christian worldview, by contrast, finds no ghosts in machines of our own making. Christianity does not exclude, however, the possibility of other intelligences, as the Bible documents encounters with angels. The consistent witness of Christianity is that human beings possess unique dignity as image-bearers of the Creator, and that Jesus Christ, the one and only Son of God, took on humanity (Philippians 2:5-7). This human dignity Christ affirmed in the Incarnation can only be seen as an obstacle to utopic projects seeking to replace human intelligence with a grand machine intelligence., The Christian, by contrast, humbly submits to the will of a loving God by seeking to have an attitude of mind like that of Christ (Romans 12:2, 1 Corinthians 2:16).
The political scientist Stephen Monsma described technology as “a distinct human cultural activity in which human beings exercise freedom and responsibility in response to God by forming and transforming the natural creation, with the aid of tools and procedures, for practical ends or purposes.” Assessments of AI that attribute moral agency to AI invert this relationship, potentially making humans the tools of machines. In reality, it is the human designers of AI who direct its programming that bear responsibility for actions of AI beyond the control of its users. C.S. Lewis anticipated this when he observed, “What we call Man’s power over Nature turns out to be a power exercised by some men over other men with Nature as its instrument.”
Projects seeking to reverse engineer the human mind in order to create a comparable or superior machine intelligence presuppose that human thought is reducible to matter and its necessary causes and random collisions. If human life is nothing more than molecules in motion, and human intelligence nothing more than neurons firing and neurochemicals churning, then there can be no place for free will, moral responsibility or ultimate purpose.
AI too easily accommodates a materialistic appraisal of intelligence. Taking as its model human intelligence, this reductionism to the physical to the exclusion of the spiritual can in turn influence how we regard our fellow human beings. We are less likely to value those whom we believe are essentially complex aggregates of macromolecules. A starkly materialistic perspective would impoverish the caring ethos essential to the moral enterprise of medicine. Oddly, some seem more fascinated with artificial than real companions. Philosopher Jay Richards observes, “The greatest delusion of our age is the paradoxical penchant to deny our own agency while attributing agency to the machines we create.”
The Christian perspective maintains that a strictly materialistic account of human intelligence is severely misguided. Human lives consist in much more than the sum of their cells; every single person is a precious soul loved by God (John 3:16-17). Accordingly, as followers of Christ we are given the ministry of loving our neighbors (Mark 12:31) and serving one another (Galatians 5:13). This calling inspires, enriches and sustains our caring roles as health professionals.
Among the promises of AI is greater efficiency. A future medical AI might, for example, read all the medical literature relevant to a patient’s diagnosis and analyze the patient’s genome prior to the appointment. The Christian healthcare professional would welcome a tool that reduces the time needed for such tasks, because this empowers the delivery of excellent medical care, as long as pressing efficiency to the maximum does not abbreviate the art of medicine. Excessive emphasis on efficiency can overlook the virtues, the character of a professional, the means by which ends are to be achieved and the special dignity and vulnerability of human lives.
If we were to grow accustomed to interacting with AIs as if they were people, we would need to take care that our habits of communication with other people did not begin to resemble automated responses or machine language. As Christians, we must also take care that our attitude toward prayer to God is not reshaped by our habits of conversation with seemingly omniscient or subservient AI devices. Our heavenly Father is the utmost source of insight and wisdom (Jeremiah 33:3, James 1:5). Unlike AI, God also grants freely to believers a sense of peace that passes all understanding (Philippians 4:7).
Further, our respect for our patient’s humanity requires our vigilance in protecting their confidentiality. We must advocate for our patient’s privacy as decisions are made regarding how AIs monitor, store and share personal medical data.
We must be aware of the capacity of AI to magnify biases in the data from which it draws and strive to prevent unjust disparities in AI conclusions and recommendations, so the poor, needy and vulnerable are not further disadvantaged (Psalm 146:5-8, James 2).
As humans endowed with free agency and moral responsibility, we must ensure the ability to override an AI healthcare decision we believe to be morally wrong or harmful to the patient. How society has understood healthcare professionals’ right of conscience in relation to written laws and policies may be reinterpreted as it confronts machine intelligence that claims access to all accumulated knowledge and is not programmed to give reasons for its decisions.
The potential medical benefits of AI are considerable and, if used wisely, can align with the Christian healthcare goals of caring for patients and avoiding harm. It matters what moral vision is embedded in AI technology. AI must remain our tool and not our master. In harnessing its power, we must not allow excessive emphasis on technical efficiency to erode the moral integrity of the medical profession or our appreciation of fellow human beings as having special dignity as bearers of God’s image.
Even if AI intelligence were to surpass human thought, Christians recognize that God’s thoughts are infinitely higher than the upper limit of anything possible by machine intelligence (Isaiah 55:8). The Christian healthcare professional understands that Christ, not technology, is the true Savior. Christ, not machine intelligence, is the way, the truth and the life (John 14:6). Christ, not artificial intelligence, is our source of wisdom, our rescue from disease and death—and our everlasting hope.
About the Author
William P. Cheshire, MD, MA, is a Professor of Neurology at Mayo Clinic in Jacksonville, Florida. Bill is a CMDA Trustee and past Chair of the CMDA Ethics Committee. He holds an AB in biochemical sciences from Princeton University, an MD from West Virginia University and an MA in bioethics from Trinity International University. His neurology residency and pain fellowship were at the University of North Carolina. This article was written prayerfully and without assistance from artificial intelligence.
This article is a written version of the presentation originally scheduled to be given by Dr. Cheshire at the 2020 CMDA National Convention. You can access more presentations from the convention, and even earn continuing education credits, through the new CMDA Learning Center at www.cmda.org/learning.
 Savulescu J. The human prejudice and the moral status of enhanced beings: what do we owe the gods?, in J. Savulescu and N. Bostrom (eds.) Human Enhancement, Oxford: Oxford University Press, 2009, p. 244.
 Waters B. From Human to Posthuman: Christian Theology and Technology in a Postmodern World. Burlington, VT: Ashgate, 2006.
 Bieber Lake C. Prophets of the Posthuman: American Fiction, Biotechnology, and the Ethics of Personhood. Notre Dame, IN: University of Notre Dame Press, 2013.
 Monsma SV. Responsible Technology: A Christian Perspective. Grand Rapids: Eerdmans, 1986, p. 19.
 Lewis CS. The Abolition of Man. New York: Macmillan, 1947, pp. 34-35.
 Cheshire WP. Till We Have Minds. Today’s Christian Doctor, Winter 2008; 39(4): 22-26.
 Richards JW. The Human Advantage: The Future of American Work in an Age of Smart Machines. New York: Crown Forum, 2018, p. 195.