The Questions AI Cannot Answer

As machines become more intelligent, humanity must decide whether our value rests on what we can do—or on who we are.

By David Rolston  |  Dogwood Creek Institute  |  On Being Human

I fell in love with artificial intelligence on a spring weekend more than fifty years ago, and I have been working in and around the field ever since.

Looking back across those decades, I am struck by how often the same questions returned: What makes a human beinghuman? Are human beings inherently valuable, and if so, why? What happens to that value if machines eventually equal or surpass our intelligence?

For many years, those questions felt mainly philosophical. Artificial general intelligence remained safely over the horizon. We could discuss the consequences without expecting to face them.

That distance is disappearing.

Why This Matters

Modern society strongly affirms that every human life has intrinsic worth. At the same time, many influential accounts of human nature describe people as entirely physical systems produced by unguided natural processes. Artificial intelligence is exposing the tension between those two convictions. If intelligence is the quality that makes us exceptional, what happens when intelligence is no longer exclusively human?

An Intuition Older Than Philosophy

For most of history, people across cultures have assumed that a human being is more than the sum of measurable parts. The language has varied—soul, spirit, sacredness, or the image of God—but the underlying intuition has been remarkably persistent. Human beings possess a dimension that cannot be reduced to physical machinery, and that dimension helps explain why human life is uniquely precious.

This intuition is older than modern philosophy and older than experimental science. It appears in the way people love, mourn, protect the vulnerable, punish murder, and insist that even a person who can contribute nothing to society still possesses dignity.

Christianity gives that intuition a specific foundation: human beings are created in the image of God. Their value does not depend on intelligence, usefulness, strength, health, age, social standing, or productivity. Human dignity is received, not achieved.

When the Old Foundation Was Removed

Over the last several centuries, the intellectual climate of the West changed. The success of modern science and engineering encouraged many thinkers to conclude that only the natural and measurable are real. Under a strictly materialist account, mind and consciousness must ultimately arise from physical processes.

A person becomes, in principle, an extraordinarily complicated biological system.  The AI community has often described this situation by simply stating that "Man is a machine."

That conclusion does not mean that materialists are uncaring or that they reject human rights. Many defend powerful secular accounts of dignity, morality, and responsibility. But a difficult question remains: Are human rights objective features of reality, or simply conventions that human societies have chosen to create and maintain?

If human value rests only on agreement, sentiment, usefulness, or intellectual ability, then its foundation may be less secure than we like to admit. Agreements change. Sentiments fade. Usefulness declines. Intelligence varies widely.

The truth is that much of modern society has, in effect, smuggled back in the conventions of the past.  Most of modern society continues to speak and act as though every person has worth that cannot be revoked. We have dismissed the old foundation intellectually while continuing to depend on it morally and emotionally.

The Personhood Paradox

This produces what I call the Personhood Paradox: We increasingly describe people as biological machines while continuing to treat them as beings with unique and immeasurable value.

Consider how differently we respond to death in nature and death among human beings. A lion kills an intelligent young elephant, and a documentary narrator describes the cycle of life. A man kills his elderly neighbor, and society holds him morally responsible, even if the neighbor was sick, dependent, and unable to contribute anything measurable.  The simple fact that we use the term "nature" to describe all of the world except humanity is a sign of how strongly we view human uniqueness.

Or compare the death of a beloved animal with the death of a child. We can grieve deeply for both, yet most people sense that the loss of the child carries a different moral weight.

Doctors devote enormous skill and resources to patients approaching the end of life, not because those patients are productive, but because their lives still matter.

Our behavior reveals a conviction that materialist theories sometimes struggle to justify: a person is never merely a unit of intelligence or economic usefulness.

“We describe people as biological machines while treating them as beings of immeasurable worth.”

The Last Refuge of Human Exceptionalism

For those who reject any transcendent source of human value, intelligence can become the last refuge of human exceptionalism. Humans matter more, the argument suggests, because we possess capacities no other known creature possesses: abstract reasoning, language, technological creativity, and broad general intelligence.

For as long as humanity has existed, we have been the only known example of general intelligence in the universe. That distinction has appeared as dependable as the sunrise.

Artificial intelligence threatens to break the connection between intelligence and humanity.

If machines display human-level reasoning, and eventually capabilities far beyond ours, intelligence alone can no longer explain why human beings deserve a privileged moral status.

The uncomfortable question then becomes unavoidable: If intelligence is what makes us valuable, should a more intelligent machine be more valuable than we are?

Gradually and Then Suddenly

These questions circulated inside the AI community for decades, but they rarely produced much fear. The possibility of machine intelligence felt like wolves howling somewhere in a distant forest. Perhaps it was more like debating the hazards of circumnavigation while relaxing aboard a sailboat still tied securely to the dock.

Then the situation changed, gradually, and then suddenly.

On November 30, 2022, ChatGPT was released to the public. The distant howling became scratching at the door.

Debates continue about whether generative AI is genuinely intelligent, whether its apparent reasoning is merely sophisticated pattern matching, and how artificial general intelligence (AGI) should be defined. Those distinctions matter. But they should not obscure what is already visible: synthetic systems now display an astonishing breadth of capability, and that capability is advancing rapidly.

The challenge does not begin only after someone announces that AGI has arrived. It begins when institutions and individuals start treating artificial systems as more capable, more reliable, or more valuable than the people they were designed to serve.

Two Very Different Foundations

If human beings have no soul, no irreducible essence, and no transcendent source of value, then human dignity must be grounded elsewhere. Perhaps it can be grounded in shared vulnerability, consciousness, relationships, moral agency, or a social commitment to mutual protection. Those possibilities deserve serious consideration. But each must answer a further question: Why should that foundation remain binding when more capable forms of intelligence appear?

If human beings bear the image of God, the answer is fundamentally different. Human value is not awarded for performance and cannot be surpassed by greater computational power. It does not depend on what we do or how much we know, but on what we are and where we come from.

Under that view, intelligence - human or artificial - is a remarkable capacity, but it is not the source of personhood. A machine may outperform us without becoming more valuable than us.

Questions Worth Considering

·       What makes a human being more than a collection of physical processes?

·       Is human dignity an objective truth, or a social agreement that future generations could revise?

·       If machines surpass human intelligence, what reason will remain for treating human beings as uniquely valuable?

·       Which qualities belong to intelligence, and which belong to personhood?

The Choice AI Places Before Us

AI may eventually force humanity to choose more explicitly between a transcendent view of the person and a purely material one. That choice could become one of the central philosophical questions of the AI age.

In one sense, the greatest challenge does not come from AI. It comes from honesty.

Machines do not need to reconcile our conflicting beliefs. We do. How long can society describe people as machines while insisting that they possess a value no machine may challenge?

What makes a human being human? Are people truly and uniquely valuable? Should we continue to treat them that way if synthetic intelligence exceeds our own?

These are questions AI can help us discuss. They are not questions AI can answer for us. The

On Being Human discussion exists to explore these questions carefully, openly, and from a Christian perspective. The conversation is long overdue. Please join us.