In the evening, in the silence of the Apostolic Palace, Pope Leo XIV opens Words with Friends. Study German on Duolingo. He challenges his brother to Wordle. He is the first pontiff who Time has included among the most influential people in the field of artificial intelligence, and this spring St. Peter’s Basilica will launch – by his will – a real-time translation system of religious services in up to 60 languages, all based on AI: just scan a QR code with your smartphone, without even downloading an app.
Having made this necessary premise, let’s move on to the interesting part. Last week, the same man summoned a group of priests to Rome to tell them of stop using ChatGPT to write homilies.
I urge you to resist the temptation to prepare homilies with artificial intelligence. Like all muscles in the body, if we don’t use them they die. The brain needs to be exercised.
Everything right, everything sacrosanct. And said with an absolutely serious face by someone who had lost to his brother at Wordle the night before.
The contradiction of the most tech Pope in history is more coherent than it seems
Here is the point that is worth understanding, because the Vatican’s position on artificial intelligence is built on a precise and overall sensible distinction. Translating the mass into Swahili or Japanese in real time is a blessed, useful, no problem tool that brings people closer to the Church.
Entrusting to an algorithm the words with which a priest should share his faith in front of his community is something different, something that belongs to the human being in such a specific way that it is impossible to delegate. The Pontifical Academy for Life has even coined a specific term for this philosophy: “algorethics”the idea that ethics should be built into the code from the beginning, sewn in, and not patched up after the damage is already done.
The Pope’s reasoning on the brain, then, also holds up outside the religious framework. Large language models like ChatGPT are very sophisticated statistical engines: they predict the most probable next word based on billions of texts. They know how to construct a grammatically correct, thematically relevant and stylistically fluent homily. They also know how to construct business emails that begin with “I hope you are well” formally perfect, emotionally empty.
A parish in the province of Foggia, the concrete concerns of those who sit on those benches on Sunday morning, the right words for those who lost someone the week before: all this remains beyond the reach of any algorithm, however sophisticated.
Don Matteo wrote to the monks to avoid social media… with a post on Facebook
The Cardinals, in the meantime, follow refresher courses on artificial intelligence, with the same resigned air as bank employees when the management system changes. Leo
Then there is Don Matteo Ferrari, from the Camaldoli monastery, in the Tuscan mountains, who wrote to his monks urging them to absolutely avoid Netflix, Instagram and TikTok, in the name of “poverty and sobriety”. He posted the letter on Facebook. No one, evidently, found the right time to tell him.
Don Cosimo Schenaa parish priest with 500 thousand followers on Instagram, said he agreed with the Pope: AI depersonalizes faith, homilies must be built on the community in front of you. He writes his, he assures, in person. Then he posts them, and half a million people read them on their phone screens. The Lord works in inscrutable ways, but with good reach metrics.
The point, ultimately, concerns everyone — priests, monks, religious influencers and anyone else with a smartphone in their hand. Do you use it to reach others, or to feel arrived? The answer changes everything, and probably not even the person concerned always knows it with certainty.
Then there is the chapter deepfakewhich partly explains the Vatican’s distrust. In 2023, Pope Francis unintentionally became the most viral spokesperson for artificial intelligence, portrayed in a designer white down jacket he had never worn. The Vatican saw something worrying in it, and published Antiqua et novaa thirteen-thousand-word document warning that AI risks “enslaving” workers and trapping children in loops of repetitive tasks. Leo XIV added his contribution, warning against digital companions “overly affectionate” who risk becoming “hidden architects of our emotional states.”
Then he reopened Wordle.
