Lecture 9-2: The Boundaries of Life — Beginning and End in the Age of AI

Psalm 139:14 — "I will give thanks to you, for I am fearfully and wonderfully made. Your works are wonderful. My soul knows that very well." (World English Bible)

Psalm 139:15 — "My frame wasn't hidden from you, when I was made in secret, woven together in the depths of the earth." (WEB)

Psalm 139:16 — "Your eyes saw my body. In your book they were all written, the days that were ordained for me, when as yet there were none of them." (WEB)

Jeremiah 1:5 — "Before I formed you in the belly, I knew you. Before you came out of the womb, I sanctified you. I have appointed you a prophet to the nations." (WEB)

Romans 14:8 — "For if we live, we live to the Lord. Or if we die, we die to the Lord. If therefore we live or die, we are the Lord's." (WEB)

3.1 The beginning of life: screening algorithms and the weight of a flag

Prenatal genetic screening has existed for decades in forms like amniocentesis and chorionic villus sampling, but AI has changed both its reach and its presentation. Machine-learning models now analyze non-invasive blood samples (cell-free fetal DNA circulating in maternal blood) and increasingly sophisticated ultrasound imaging to estimate the statistical likelihood of chromosomal conditions such as Down syndrome (trisomy 21), along with a growing list of other genetic markers. These tools are genuinely useful for some purposes — identifying pregnancies that may benefit from closer monitoring, or preparing parents in advance for a child's medical needs — and this lecture does not treat the existence of such screening as itself the ethical problem.

The entanglement this lecture is concerned with is specific: in a substantial number of documented cases internationally, a positive screen for a condition like Down syndrome is followed, at a very high rate, by a decision to end the pregnancy — a pattern well-documented enough that some countries with widespread prenatal screening report Down syndrome births falling sharply as screening uptake rises. The algorithm itself does not make this decision; it produces a probability. But the probability is delivered into a context — often a brief clinical encounter, often under time pressure, often with parents unprepared for the news — in which a statistical flag can function, in practice, much more like a verdict than a probability, especially when it is presented (as these tools often are marketed) with an emphasis on its precision rather than its uncertainty or its silence on what a life with the flagged condition is actually like to live.

This is where the lecture's theological argument bears most directly. Psalm 139 describes the psalmist's own gestation in language that is deliberately intimate and pre-cognitive: "my frame wasn't hidden from you, when I was made in secret, woven together in the depths of the earth" (v. 15, WEB), and "your eyes saw my body. In your book they were all written, the days that were ordained for me, when as yet there were none of them" (v. 16, WEB). The psalm places God's attention and God's writing of a life's days at a stage prior to any measurable capacity, prior even to the body being formed in a way anyone else could observe — "in secret," in the psalm's own phrase. Jeremiah 1:5 makes a structurally identical claim about vocation rather than dignity in general: "before I formed you in the belly, I knew you... I have appointed you a prophet to the nations" (WEB), grounding a specific calling in a divine knowledge that precedes gestation itself.

Neither text was written to settle a modern screening-and-selection debate, and this lecture does not claim they were. What they establish, read on their own terms, is that the biblical pattern for locating a life's worth is never "wait until a measurable capacity appears" — it is the reverse: dignity and calling are affirmed prior to and independent of anything measurable, at the very stage a screening algorithm's statistical confidence is highest and a life's observable capacities are at their lowest. Whatever ethical work remains to be done in a specific hard case, this is not a small foundation to bring to it: a Christian bioethics of the beginning of life resists letting a probability about a genetic marker function as though it disclosed anything at all about the worth of the life it describes.

3.2 The end of life: prognosis tools and the temptation of false finality

AI-assisted prognosis tools serve a parallel function at the other edge of life. These systems analyze a patient's clinical data — vital signs, lab values, comorbidities, sometimes imaging — to predict outcomes such as likelihood of survival to discharge, likelihood of meaningful neurological recovery after a severe brain injury, or expected mortality within a given time window. Hospitals increasingly use such tools to inform, though rarely to dictate outright, conversations with families about whether to continue or withdraw life-sustaining treatment.

The genuine value of such tools should not be dismissed: families and physicians facing an end-of-life decision often want more information, not less, and a well-validated prediction can be one honest input among several. The danger this lecture names is not the existence of prognostic information but its false finality — the tendency, under the pressure and grief of the moment, for a percentage ("twelve percent chance of meaningful recovery") to function as though it settled a question it cannot settle: whether continued life, at this probability of this quality, remains worth sustaining. A prognosis is a prediction about the body's likely trajectory; it is not, and does not become, a statement about the worth of the life whose trajectory it predicts, even though the two are easily confused precisely because the number arrives dressed in the same statistical authority medicine uses for genuinely settled clinical facts.

Romans 14:8 supplies this lecture's governing text for resisting that confusion: "For if we live, we live to the Lord. Or if we die, we die to the Lord. If therefore we live or die, we are the Lord's" (WEB). Paul's immediate context is a dispute within the Roman church about matters of conscience — food, sacred days — remarkably far from a hospital's intensive care unit. But the verse's underlying claim is precisely on point for this lecture's purpose: a person's belonging to the Lord, in life and in death, is not contingent on a measurable quality-of-life threshold or a survival probability crossing some line. This does not settle, by itself, whether continuing or withdrawing a specific treatment in a specific case is the right decision — those decisions properly weigh many factors, including a patient's own previously expressed wishes, the burdens a treatment imposes, and reasonable medical judgment about futility, none of which Romans 14:8 addresses directly. What the verse forecloses is treating a prognosis algorithm's output as though it were itself the final word on whether a life remains worth living — a category error the text's theology does not permit, whatever the actual clinical decision turns out to be.

3.3 Being honest about disagreement

Christians who fully affirm both arguments above — that dignity precedes measurable capacity, and that a prognosis is not a verdict on a life's worth — still disagree, sometimes sharply, about specific applications. Some hold that withdrawing artificial nutrition and hydration from a patient in a persistent vegetative state is a legitimate recognition that a particular intervention has become disproportionately burdensome relative to its benefit, not an act of ending a life whose worth has been judged insufficient. Others hold that this same act crosses a line between allowing death and causing it, regardless of how the decision is framed, because withdrawing sustenance (as opposed to withdrawing, say, a ventilator) removes something more basic than medical treatment. Catholic moral theology's distinction between "ordinary" and "extraordinary" means has been a long-standing and serious attempt to navigate exactly this line, and it does not produce unanimous verdicts even among those who accept the distinction's basic framework, because reasonable people disagree about which specific interventions, for which specific patients, count as which.

This lecture takes no position on which side of that specific dispute is correct, and it would misrepresent the actual state of careful Christian bioethical reflection to imply otherwise. What it does insist on is that this disagreement occurs entirely downstream of the two shared commitments argued for above. No side in this debate holds that a prognosis algorithm's number should function as the actual decision-maker, and no side holds that dignity is contingent on a measurable threshold. The disagreement is about how those shared commitments cash out in specific, genuinely hard clinical circumstances — which is a very different, and more honest, description of the state of the debate than a simple story in which one side "trusts God" and the other does not.