Tonight, at the opening Mass of the DELTA Summit on AI, Faith and Human Flourishing at the University of Notre Dame, Cardinal Blase Cupich asked a question that has become increasingly difficult to avoid: What does it mean to be human in the age of artificial intelligence?
But Cupich did something unexpected with the question. Rather than beginning with what artificial intelligence might someday become, he turned the question back on us.
The tendencies we rightly fear in AI—the reduction of people to data, the sorting of human beings into categories, the prediction of behavior, the measuring of worth according to usefulness or productivity—were present long before anyone trained a large language model.
“AI is holding up a mirror to us,” Cupich said, revealing our own tendency to see and judge one another in “cold and deterministic ways.”
That may be one of the most important theological observations we can make about artificial intelligence.
Before the algorithm, there was the human heart
Preaching on the call of Matthew, Cupich pointed to a man everyone around him thought they already understood.
Rome saw a functionary. His neighbors saw a traitor. The religious establishment saw a sinner. His occupation had become his identity, his past had determined his future, and his worth could seemingly be entered on a ledger.
Cupich memorably called Matthew “the ancestral prototype of the modern digital citizen.” Everyone had enough data to construct a profile of him.
Then Jesus walked by.
Where everyone else saw a category, Jesus saw a person.
Where society had reduced Matthew to his usefulness, failures and history, Jesus saw someone made in the image of God. Christ’s gaze did not merely describe Matthew more accurately. It opened a future that the categories imposed upon him could not imagine.
That distinction matters enormously in our conversations about AI. The danger is not simply that machines will learn to reduce human beings to data. We already know how to do that.
We do it whenever productivity becomes the measure of a worker’s worth, whenever someone’s worst mistake becomes the permanent definition of who they are, and whenever what can be measured about a person becomes a substitute for knowing the person.
Artificial intelligence can perform these reductions at extraordinary speed and scale. But the temptation behind them is profoundly human.
A person is more than a prediction
This leads Cupich to what I think is the central claim of the homily: the challenge posed by AI “is not merely a technological question of algorithmic design, but an anthropological one.”
Before asking what kind of artificial intelligence we want to build, then, we need to ask what we believe a human being actually is.
That question cannot finally be answered by comparing human capacities with machine capacities. Machines will become better at many things we once regarded as distinctively human. They can already calculate faster, remember more, recognize patterns invisible to us and imitate increasingly sophisticated forms of human communication.
Christian anthropology begins somewhere else.
Our dignity is not something we achieve. It is something we receive.
Cupich says our fundamental identity “is already given to us” and “is not forfeited by our weakness.” Then comes the heart of his answer:
“It is love.”
Any account of human dignity built upon intelligence, productivity, independence or usefulness eventually encounters people who cannot meet the standard: the infant, the person with profound intellectual disability, the person whose dementia has carried away memory and language, the dying person who can no longer produce or perhaps even respond.
None has become less human.
Their dignity was never theirs because of what they could do.
What kind of people are we becoming?
This is where tonight’s homily raises the question I hope follows the DELTA Summit through the coming days.
Cupich warns that AI’s deterministic logic could become “the template for how we see ourselves and our neighbors.”
Technology does not simply do things for us. It also forms us.
The tools we repeatedly use shape our habits of attention and our expectations of one another. They can encourage patience or impatience, responsibility or avoidance, relationship or transaction. As more of our lives are mediated by systems designed to predict what we will want, say, buy, watch and do next, we should ask what happens to our ability to encounter another person as mystery.
Because a person is not a prediction.
A person always exceeds the profile.
There is something wonderfully subversive, then, about Jesus walking past Matthew’s customs post. Everyone else thinks they already know who Matthew is.
Jesus sees who Matthew might yet become.
The mirror
Perhaps that is why Cupich’s image of AI as a mirror is so useful. A mirror does not create the face it reveals.
If artificial intelligence reflects our obsession with efficiency, categorization, productivity and prediction, the answer cannot simply be better engineering. Better engineering matters. Ethical safeguards matter. Regulation matters.
But none can substitute for conversion.
Cupich closes with Gerard Manley Hopkins and the reminder that beneath all the grime humanity manages to smear across creation, “there lives the dearest freshness deep down things.”
Perhaps that is one of the unexpected gifts of this technological moment. AI is forcing us to look again at assumptions about intelligence, agency, productivity, relationship and worth that were never created by machines in the first place.
The machine may hold up the mirror.
What we do with what we see remains our responsibility.
And perhaps the most urgent question in the age of artificial intelligence is not whether machines will eventually become more like us.
It is what kind of human beings we are becoming in their presence.
