”Magnifica Humanitas.” What Unites the Mathematician Pope to the AI Technocrats. And What Divides Them

The experts who, along­si­de the pope, on Monday, May 25, will pre­sent to the world “Magnifica Humanitas,” Leo XIV’s fir­st ency­cli­cal, fea­tu­re the English theo­lo­gian Anna Rowlands of Durham University, reci­pient in 2023 of a Joseph Ratzinger-Benedict XVI Foundation award, and the American entre­pre­neur Christopher Olah (pic­tu­red), co-founder of Anthropic, both among tho­se most com­mit­ted to addres­sing the cru­cial issues rai­sed by arti­fi­cial intel­li­gen­ce, to which the ency­cli­cal is dedi­ca­ted.

Signed by Leo on May 15, exac­tly 135 years after his pre­de­ces­sor and name­sa­ke Leo XIII signed the fir­st, histo­ric ency­cli­cal “Rerum nova­rum” dedi­ca­ted to the social doc­tri­ne of the Church, “Magnifica huma­ni­tas” also aims to respond to the exi­sten­tial que­stions posed by the cur­rent new revo­lu­tion in human socie­ty : the one crea­ted pre­ci­se­ly by arti­fi­cial intel­li­gen­ce.

Anthropic is not the only big com­pa­ny acti­ve in this field. One need only think of Alexander Karp and Peter Thiel’s Palantir, Sam Altman’s OpenAI, Elon Musk’s xAI and Grok, each con­vey­ing a dif­fe­rent techno-philosophical vision.

Thiel made a stir with his appea­ran­ce in Rome last March for a series of closed-door con­fe­ren­ces on the Antichrist. But more than his apo­ca­lyp­tic vision inspi­red by René Girard, what car­ries weight on the poli­ti­cal ter­rain is his pro­xi­mi­ty to JD Vance, the vice pre­si­dent of the United States, a con­vert to a Catholicism of rup­tu­re with the pre­vai­ling trends of the Church and a scorn­ful cri­tic of a Europe that, with its AI Act pas­sed in 2024, pre­su­mes to regu­la­te arti­fi­cial intel­li­gen­ce throu­gh legi­sla­tion, clas­si­fy­ing and sanc­tio­ning its risks pre­emp­ti­ve­ly – an illu­so­ry under­ta­king in a con­stan­tly evol­ving field.

Anthropic, instead, is the expres­sion of a highly ori­gi­nal vision to which the Church of Rome pays clo­se atten­tion. And that is pre­ci­se­ly why Pope Leo cal­led on Olah to pre­sent “Magnifica huma­ni­tas.”

To bet­ter under­stand this vision, it is worth repea­ting word for word the descrip­tion of it given in the May 18 edi­tion of "Il Foglio" by a great expert on the sub­ject, Carlo Alberto Carnevale Maffè, pro­fes­sor of busi­ness stra­te­gy at Bocconi University in Milan and brought in to teach at some of the most pre­sti­gious uni­ver­si­ties in the world, from Columbia University to the Wharton School, from Steinbeis University in Berlin to St. Mary’s College of California.

Together with Olah, the Anthropic co-founders inclu­de Dario Amodei, cur­ren­tly its chief exe­cu­ti­ve offi­cer, and his sister Daniela. And it is the essay “Machines of Loving Grace,” which the two publi­shed in 2024, that best expres­ses their vision, which is also poli­ti­cal.

“It is fif­teen thou­sand words worth rea­ding in their enti­re­ty,” wri­tes Carnevale Maffè, “befo­re expres­sing any judg­ment on Silicon Valley. Their the­sis is clear : ‘We see no struc­tu­ral rea­son why AI should pre­fe­ren­tial­ly favor demo­cra­cy and pea­ce.’ It is an admis­sion that none of their col­lea­gues has had the cou­ra­ge to for­mu­la­te with such cla­ri­ty, and that on its own would merit a poli­ti­cal phi­lo­so­phy semi­nar.”

“Amodei reco­gni­zes,” Carnevale Maffè con­ti­nues, “that AI can enhan­ce pro­pa­gan­da and sur­veil­lan­ce, the two clas­sic tools of the auto­crat, and that demo­cra­cies must the­re­fo­re play acti­ve­ly to gain a struc­tu­ral advan­ta­ge, not being able to tru­st in tech­no­lo­gi­cal iner­tia. It is a posi­tion that sepa­ra­tes Amodei from the opti­mi­stic deter­mi­ni­sm that domi­na­ted Californian thought in the nine­ties : the idea, that is, of vague­ly Clintonian deri­va­tion, that the inter­net would auto­ma­ti­cal­ly export demo­cra­cy (we all remem­ber the ‘arab springs’ and the illu­sions that fol­lo­wed them). Amodei expli­ci­tly demo­li­shes that nar­ra­ti­ve : ‘The inter­net has pro­ba­bly bene­fi­ted autho­ri­ta­ria­ni­sm, not demo­cra­cy.’ This is an impor­tant and sur­pri­sing histo­ri­cal cor­rec­tion for an American CEO in this sec­tor.”

Hence Amodei’s ope­ra­tio­nal pro­po­sal, what he calls an
“enten­te stra­te­gy.” “It is a coa­li­tion of demo­cra­cies that would ensu­re pri­ma­cy in AI throu­gh con­trol of the chip sup­ply chain and stra­te­gic mili­ta­ry action (‘the stick’) com­bi­ned with the distri­bu­tion of bene­fi­ts (‘the car­rot’) to shift the glo­bal balan­ce.”

In a sub­se­quent essay from 2025, “The Adolescence of Technology,” Amodei expan­ded on this line, “adding an unea­se that has beco­me his theo­re­ti­cal tra­de­mark,” Carnevale Maffè fur­ther wri­tes. “The risk he warns again­st is not only that auto­cra­ts will use AI again­st demo­cra­cies, but that demo­cra­cies them­sel­ves, in the name of effi­cien­cy, will sli­de toward forms of dome­stic techno-authoritarianism. The ‘coun­try of geniu­ses in a data cen­ter,’ an Amodeian phra­se that has now ente­red the com­mon lexi­con, is a con­di­tio­nal uto­pia : it is only valid if the insti­tu­tio­nal geo­me­tries can with­stand the blow of the con­cen­tra­tion of com­pu­ta­tio­nal power.”

Of all the posi­tions in the field, Carnevale Maffè com­men­ts, “Anthropic’s is the most reco­gni­za­bly Kantian in form and Churchillian in sub­stan­ce. It is no coin­ci­den­ce that it is also the most respec­ted in Western aca­de­mic cir­cles and the only one, it must be said, that has taken the trou­ble to seek cri­ti­cal com­men­ta­ry, promp­ting public deba­tes like that at Cambridge’s Leverhulme Centre for the Future of Intelligence, which pro­du­ced a stern but con­struc­ti­ve rea­ding of Amodei’s essay.”

Anthropic’s co-founders are not the only ones acting on the basis of a techno-philosophical vision. Alexander Karp, CEO of Palantir, holds a PhD in social theo­ry from the University of Frankfurt, and in his 2025 essay, co-authored with Nicholas Zamiska, “The Technological Republic : Hard Power, Soft Belief, and the Future of the West,” he wri­tes with the tone of a for­mer stu­dent of Jurgen Habermas and the Frankfurt School. Translated into prac­ti­ce, his the­sis is that the West needs to build an AI-industrial com­plex ana­lo­gous to the military-industrial com­plex of the Eisenhower era if it wan­ts to with­stand cogni­ti­ve com­pe­ti­tion with auto­cra­tic regi­mes.

But if Karp, with Palantir, wan­ts to main­tain and indeed streng­then the histo­ri­cal col­la­bo­ra­tion with the American govern­ment, this is not the case with Olah and Amodei, who­se com­pa­ny, Anthropic, was ban­ned by Donald Trump last February pre­ci­se­ly becau­se of its refu­sal to allow the United States mili­ta­ry unli­mi­ted use of its AI tech­no­lo­gy.

It is no sur­pri­se, then, that Pope Leo, alrea­dy at odds with the White House, should have wan­ted none other than Olah to pre­sent “Magnifica huma­ni­tas.” There is an unde­nia­ble con­so­nan­ce, on the sub­ject of arti­fi­cial intel­li­gen­ce, bet­ween the vision of Anthropic’s co-founders and that of the Church of Rome, alrea­dy discer­ni­ble in that pre­view of the new ency­cli­cal which was the docu­ment “Antiqua et Nova,” publi­shed by the dica­ste­ry for the doc­tri­ne of the faith in January 2025.

New tech­no­lo­gi­cal pro­duc­ts, in fact, as one reads in “Antiqua et Nova,” are not neu­tral : they “reflect the world­view of their deve­lo­pers, owners, users, and regu­la­tors, and have the power to sha­pe the world and enga­ge con­scien­ces on the level of values.” And this, Carnevale Maffè obser­ves, “is exac­tly the same cri­ti­ci­sm that Habermas and the Frankfurt School would have made.” Leo XIV, the mathe­ma­ti­cian pope from Villanova University, “is not play­ing again­st Silicon Valley. He is play­ing with intel­li­gent Silicon Valley again­st its more cru­de, chau­vi­ni­stic, and ido­la­trous ver­sion.”

In other words : “If one accep­ts this car­to­gra­phy, the distan­ce bet­ween Leo’s Augustinian per­so­na­li­sm and Anthropic’s cau­tious demo­cra­ti­sm is, in terms of goals, much smal­ler than the distan­ce that sepa­ra­tes both from Vance’s Trumpism and Musk’s liber­ta­ria­ni­sm.”

But then the­re are also the visions of Karp and Thiel, more que­stio­na­ble but not to be igno­red, with a view to a cri­ti­que of autho­ri­ta­rian tech­no­cra­cy to be car­ried out in allian­ce with the heal­thy part of techno-politics.

“This is what the Church has always been able to do when it has wor­ked well,” Carnevale Maffè con­clu­des. “Keeping toge­ther Thomists and Franciscans, Jesuits and Dominicans, in the name of a truth grea­ter than all schools. Different means, dif­fe­rent litur­gies, dif­fe­rent cathe­drals : Karp’s data cen­ter and St. Peter’s Basilica. But the ene­my is the same. And History, when it wan­ts to be mali­cious, backs the most sur­pri­sing allian­ces into the most unex­pec­ted cor­ners.”

(Translated by Matthew Sherry : traduttore@​hotmail.​com)

— —  — —

Sandro Magister is past “vati­ca­ni­sta” of the Italian wee­kly L’Espresso.
The late­st arti­cles in English of his blog Settimo Cielo are on this page.
But the full archi­ve of Settimo Cielo in English, from 2017 to today, is acces­si­ble.
As is the com­ple­te index of the blog www.chiesa, which pre­ce­ded it.

Retour en haut