Written by Professor Edward Harcourt MBE
“What a piece of work is man,” says Hamlet. Nowadays people are more likely to marvel at machines, especially machines driven by AI. Just think of the robot half marathon in Beijing a few weeks ago – no sweat, no pride or exhaustion in its expressionless face of the ‘winner’ as it tick-tocked over the line at its mindless, metronomic pace, but the crowds cheered all the same. But as of 15 May and the publication of Pope Leo XIV’s encyclical on AI, humanity has a new, powerful voice speaking up for us. The title says it all: Magnifica humanitas – magnificent humanity. Hamlet would surely have approved.
Whether Hamlet would have approved or not, there is much wisdom in what the Pope has to say. He favours protecting the dignity of work. He urges clarity on liability and he’s right: one of the biggest if not the biggest AI governance challenge is where to assign liability when things go wrong. He favours “the need to strengthen each person’s interior freedom.” Too right: nurturing the soft infrastructures that give us a critical distance on AI-generated advice are key to staving off cognitive atrophy, the condition in which our powers of cognition and practical judgment wither because we defer too much to AI, and wither even if we keep an uncritical “human in the loop.”
All that said, the Pope is not – forgive the expression – infallible. He calls for an ethical code for AI subject to shared standards of social justice, because “a more moral AI is not enough if that morality is determined by a few.” But the call for a “code” – whether determined by few or by many — builds on an assumption common (for example) to Anthropic’s “Constitution” for Claude: that ethical AI is all about training AI to think ethically, and that that involves seeking consensus.
But mightn’t people reach a consensus on something bad, or trivial? And mightn’t pluralism about the best way to live be a good thing in its own right? Some of the most promising thinking about AI ethics isn’t about devising a code for AI at all, but about organising the AI ecosystem in such a way that no one voice or set of voices drown others out – the family of infrastructural innovations known as “decentralized AI” for example. As Magnifica Humanitas itself says, we should oppose the overconcentration of technological power in the hands of a few.
Be all that as it may, what stands out about the encyclical is not its particular recommendations but, underlying them all, its affirmation of the “grandeur of humanity” – what Hamlet saw clearly but what we, dazzled by technology, are in danger of losing sight of. The very phrase “artificial intelligence” encodes the assumption that there’s a single feature – intelligence – that is possessed both by us and by AI. That’s questionable: Nick Bostrom’s influential book Superintelligence imagined an AI-powered robot that destroys humanity in its zeal to carry out an instruction to manufacture paperclips. But a robot that did that could not be called intelligent. In fact, it would be super stupid: it would have all the intelligence of the recovering alcoholic who jailbreaks the rehab centre to make it to the off licence. That might be what Aristotle calls “cleverness,” but it’s not real intelligence, because it lacks the power of judgment to assess what’s worth doing, and for the sake of what.
Of course, one might argue that the issue is not what AI can do, but what AI could do if the tech develops further. Intelligence is what the experts call “‘substrate neutral” — we may only recognise it now in humans, but the human brain is not the only place it can be encoded. Sufficiently advanced silicon chips, the thought goes, could accommodate intelligence just fine.
But if the Pope is on to anything, that misses the point: it is the very humanness of human beings, not what we “encode,” that makes all the difference. That should be evident from the fact that we matter to one another in ways that have nothing to do with our cognitive powers. As Simone Weil says, “[N]o one stands up, or moves about, or sits down again in quite the same fashion when they are alone in a room as when they have a visitor.” A human being with dementia is a tragedy, not a technical failure. A dead human being is a very powerful presence; a robot that has stopped working is a piece of rubbish.
But reassuring as it may seem for humanity to have acquired such a prominent and influential ally as Pope Leo, I can already hear the rumbles of doubt and disapproval. Roman Catholic teaching on many subjects – abortion, homosexuality – is controversial, so if they stress the special dignity of the human person, can people who are not members of that faith community really trust their opinion? And anyway – again, if they say it – the special dignity of the human person must have a theological foundation: humanity, as Pope Leo says, is created in the image of God. And don’t the people who say that also say that the world was created in seven days, and all that? Mightn’t this “humans are special” thing be bound up with the idea of an immaterial soul? But then the idea that humanity is uniquely magnificent must be a niche commitment of people who have swallowed a whole lot of scientifically suspect theology as well.
What’s more, if we haven’t swallowed all that, the worry is that we’re stuck with the cultural mainstream view that goes all the way back to Hobbes and Descartes: that humans are machines. And that’s risky: if humans and AI are just two kinds of machine, what place in the world do we deserve if AI is a more powerful machine than we are?
It’s a sad state of affairs if there’s a binary choice between Roman Catholic doctrine on the one hand and the anti-humanism of the cultural mainstream on the other. But there isn’t. Years ago philosopher Cora Diamond published a paper called “Eating Meat and Eating People.” Her point was not that it’s OK to eat meat but not OK to eat people: as a matter of fact she’s a vegetarian. It’s rather that our reasons not to eat people are of an entirely different order.
If the concern was just that it’s wrong to kill people for food, eating people who are already dead would be fine. But nobody thinks that. The fact that we don’t see people as a food source at all is just one aspect of our specialness to one another. Another is the fact that we give human beings names rather than numbers, and not just any old names but names that locate us in further networks of significance such as ancestry or wider group membership. Or that we see even dead human bodies as needing to be treated with special solemnity, even though intelligence has vanished from them; or that we bring to everything we do a piece of knowledge that generally stays in the background - that we have all been born, and are going to die.
Our specialness to one another resides in facts like these – “modes of response,” as Diamond calls them — rather than in neatly articulated codes of belief or “values”. But for that very reason - because it is so often habitual, inexplicit, mediated by our vulnerable bodies and by our infinitely mobile faces in which we can read, or fail to read, second by second changes of mind – the varieties of this specialness can often be invisible to us. Just as we can thrive on fresh air without knowing the chemical formula for it, it takes intellectual effort to piece together these “modes of response” and bring them to explicit awareness. And because the implicit is a poor match for the explicit, that is also why the recognition of our specialness to one another is vulnerable to hostile ideologies – including those that are associated with AI – and why today’s encyclical is such a helpful antidote.
But in the context of Magnifica Humanitas, the crucial point is that none of these facts about us – none of these “modes of response” — assumes any kind of theological foundation. On the contrary, our specialness to one another is hard-wired, part of our biological endowment as creatures of a certain species. Rival ethical theories, rival catalogues of virtues that point towards different visions of the well-lived life and indeed different faith traditions, including the idea that we are made in the image of God, express our differing attempts to give a more determinate shape to the specialness to one another of which we’re constantly but inexplicitly aware in our daily embodied interactions with one another.
Some of these may be correct and others not, and I am not arguing for or against any one of them. I’m arguing that recognising the special dignity of human beings, as the Pope has rightly urged us to do, doesn’t mean we have to buy into his foundations. It’s not the theology – or indeed this or that ethical theory – that makes for the specialness. Our specialness to one another comes first.
Suggested citation: Professor Edward Harcourt MBE, ‘Magnificent Humanity’, (12 June 2026), AI Ethics at Oxford Blog; https://www.oxford-aiethics.ox.ac.uk/blog/three-cheers-magnifica-humanitas-uniqueness-human-doesnt-need-theological-foundation