AI and the Dignity of the Person

I am not a sceptic. I am curious beyond repair — and the scars on my back are the tuition I paid for a single lesson: in the end, only so much matters.

Four decades of trying to understand what truly drives company performance teach you to tell the vital few things from the many that merely fill the conversation. I carried that instinct into the Oxford Saïd programme on artificial intelligence, and it came out sharpened, pointed at something new.

First, the demystification. Much of what is sold as machine intelligence is, mechanically, pattern recognition and next-token prediction — a remarkably good guess at what comes next, run at extraordinary scale. That is not a dismissal; it is a clarification. Strip away both the magic and the fear, and you can ask the only question worth asking: where does this actually create value, and where is it theatre?

Here the oldest lesson in my trade proved the most useful. Value in any organisation concentrates. It does not spread evenly across roles and initiatives; it pools in a few places that disproportionately decide the outcome. I have spent a career arguing that talent should be pointed deliberately at those points rather than sprayed uniformly. Oxford showed me that AI obeys the same law. It compounds where it is aimed at a genuine point of leverage, and dissipates into cost where it is adopted merely to look modern. The technology has not changed the question.

The winners will not be those running the most experiments, but those with the discipline to end the ninety that do not matter and resource the ten that do.

Which makes the companion discipline decisive: subtraction. The strategic act is naming what to stop, not what to add. Most AI strategies I see are the reverse — a proliferation of pilots, each defensible, together an exercise in overload.

For those of us in leadership and in HR, this is not a threat to our work; it is a return to its point. AI does not replace the questions we were always paid to answer — which roles truly move the needle, how work is designed, whether managers are good enough to carry the strategy. It sharpens them. A powerful tool in an undisciplined organisation simply industrialises the waste.

And then the part that moved me most. Pope Leo XIV, in his encyclical Magnifica Humanitas, opens and closes where everything human should: with the dignity of the person. He asks of artificial intelligence the same question the Church put to industrial capitalism in 1891 — does it treat a person as a person, or as a cost? I was raised a Catholic, and though I no longer hold the belief, I remain bound to the spirituality of the church, and I hear in his argument something truer than any productivity forecast. The economist Daron Acemoglu arrives at the same place by another road: AI's real value will be narrower, slower, and wholly dependent on how — and on whom — it is deployed. Both are saying what good practitioners have always known. Tools do not create value. Judgement about where to point them does — and the finest judgement keeps the human person, not the machine, at the centre.

That judgement was always the job. AI has only made it more consequential to get right.