I have long had a special admiration for the French mathematician and religious thinker Blaise Pascal (1623-1662), who understood better than any of the other pioneers of modern science and mathematics that the closure of knowledge was neither to be desired or feared, and who foresaw that the exploration of nature and the ever-more-subtle tools required to understand it would not, as Enlightenment rationalists and their progeny have continued to think, lead to a “theory of everything” that would exhaust the realm of knowledge of the natural world. Whence this third in a series of Chronicles paying homage to this classical master of paradox as the first true precursor—in spirit if not in detail—of Generative Anthropology.
In this light, the evolution of my own thinking from The Origin of Language until today can be seen as an effort to reconcile the two domains of what Pascal called géométrie and finesse; the first, easily enough translated; the second, the home of the un-geometrical, what we might call the “intuitive,” but I think best understood as the intrinsically paradoxical realm of the sacred, which was most certainly, beyond the realm of geometry and its fellow sciences, this pious Jansenist’s principal preoccupation.
On March 31, 2025, the conservative French daily Le Figaro published what we would call an op-ed by Laurent Alexandre, Olivier Babeau, and Alexandre Tsicopoulos entitled “La grande démission de l’intellect face à l’IA”: “The Great Abdication of the Intellect in the Face of AI.”
After citing a study showing that 88% (!) of today’s British upper-school students write their papers with the help of ChatGPT, the authors point to the steady decline of European Pisa test scores. Thus “one already observes in certain professions a phenomenon of de-qualification: the intensive use of AI leads to the loss of certain skills that are no longer activated. Imagine what will be the case in the future when these skills will not even have been acquired.” And in conclusion, “the future that is emerging is closer to the plot of the 2006 film Idiocracy than to the flourishing of an advanced society.” (Idiocracy was a wonderful comedy in which two 21st-century humans were placed in suspended animation in a time capsule and woke up a few centuries later to discover themselves in a world where the machines did all the thinking, while the humans had become… what the film’s title suggests.)
There is no doubt that the Figaro authors are posing a serious problem that will have to be faced up to by a carefully programmed modification of the tools of instruction. Having had the good fortune to retire before this latest disastrous innovation—following the already-disturbing move to online teaching occasioned by the pandemic of 2020—I can only commiserate with those now obliged to assign papers to be written in class (without cell phones or laptops), and shortly perhaps to pass their students through a detector on entering the classroom to prevent them from exchanging clandestine electronic signals.
But the real point of this Chronicle remains nonetheless Pascalian. AI can surely defeat human intelligence in any algorithmic task, and its latest developments extend this ability into areas far broader than mathematical calculations or even the game of Go. But Pascal’s esprit de finesse, like his gambler’s decision to bet on God’s existence, is a product of a transcendence that cannot be reproduced in a machine, unless machines could be made to reproduce the very history embodied quite literally in our human flesh.
The esprit de finesse depends on what we call insight, the discovery of a connection not previously made, and of which the esprit de géométrie is in principle incapable, because insight cannot be derived from the axiomatic system on which the latter depends for its “intelligence.”
And this is true as much for the linguistic “intelligence” displayed by ChatGPT and its competitors as for the strictly mathematical calculations of traditional computers. The “wisdom” of interactive Large Language Model programs of this kind derives from recombining presumably reliable assertions from online reference material. The only connections they can make between statements must already be implicit in the vocabularies of the statements themselves; they have no capacity to discover new connections not already at least hinted at in their data bases. Whereas the insights of the esprit de finesse make use of human beings’ experiences not yet formulated as factual statements. Observing reality can lead me to discover new facts about reality in a way that observing previous statements about reality cannot.
Given that what we have already learned about the material world suggests the truth of Pascal’s intuition of the two infinities: that we have not reached—and will quite likely never reach—the limit of smallness or largeness, either in time or in space, so that no algorithmic system can replace our continually improved observation of the natural world, which itself can never be definitively summed up by a set of equations. And whether or not there are elsewhere creatures other than ourselves capable of constructing scenes on which to model the real world through representations, this process is not reproducible within the cybernetic realm in its current state.
So that for the moment at least—and this “moment” may well be either longer or briefer than we imagine—the esprit de finesse that is the home not merely of “intuition” but of what we humans call the sacred must still be cultivated. All the more so given the danger pointed to by the Figaro article: that in losing our abilities to perform algorithmic acts—and many/most? people today can no longer multiply, much less divide, two multi-digit numbers with pencil and paper—we will dull our mental acuity and our sensitivity to the insights of finesse as well.
The sacred and the infinities
Reading Melanie Phillips’ critique of postmodern thinking in The World Turned Upside Down (Encounter, 2010), I certainly sympathize with her rejection of the dogmatic atheism of such as Richard Dawkins and Steven Pinker. Thinkers who dismiss religion as nonsense turn their backs on a foundational element of the human itself, as though it belonged to the same category as empirical misapprehensions. On the contrary, Roy Rappaport’s great intuition that language and religion were coeval at the birth of what we can truly call the human obliges us to treat the phenomenon of the sacred as an inseparable component of it, independently of whatever specific beliefs in sacred beings the different creeds attach to it. What is indeed paradoxical is that a religion that seeks to embody the sacred independently of all such beliefs is effectively inconceivable. To understand the sacred “in itself” within our minimal hypothesis, we can only posit its behavioral origin in the originary scene of language; its historical developments in the form of collective religions then belong to the history of human culture. The question of evaluating religious beliefs as historical facts is a very different matter from evaluating the functioning of these beliefs within the human societies that have adopted them.
No doubt to take this position forces us to reject the usual apologies for religion in its own terms, which in effect oblige us to accept as true without proof assertions about reality that we can only know as those of our fellow humans. But the task of anthropology is to grasp in what way religion, as the institutional organization of the sacred, operates in specific cases to maintain the human communities that have adopted them.
In the scenario of the originary hypothesis, the sacred originates as our internal sense of the interdiction, whose locus comes to be called our conscience, of potentially conflict-producing “instinctual” appetitive acts in the interest of maintaining order in the human community. I have given as a minimal (but not originary) expression of this interdiction the example of Hillel’s “golden rule”: do not do to another what you would not have him do to you.
GA situates the originary experience of the sacred in the hesitation that, ex hypothesi for the first time, obliges the individuals surrounding an object of appetitive desire, such as the body of a prey animal, to defer seeking to appropriate this object for themselves until such time as all can agree to share it. It is at this point that the “conditioned-reflex” inhibitions that prevent conflict among animals may be presumed to have become incapable of regulating living creatures once they have attained a certain level of (mimetic) desire, and must therefore be reinforced by the conscious rejection of the appropriative act.
This deferral of appropriative action then presumably leads the community to share a pointing or ostensive gesture that constitutes the first sign of language: the first shared signifier by which the members of the group designate the object of their common desire, and in that act, qualify it as sacred, a term to which we can give the minimal meaning of belonging in the first place to the community rather than being available for possession by any given individual.
In the absence of archaeological evidence, the only demonstration of this hypothetical scenario can come from Occam’s razor. But the Derridean term différance, which its creator unfortunately demurred to understand in anthropological terms, nevertheless admirably demonstrates the link intuited by Rappaport between the births of language and religion: the sacred is in the first place what can only be referred to, pointed to by what will become in that very act a sign. And this minimal ostensive sign must be understood not simply as designating a worldly thing, let along this sort of thing, like the word “cat,” but rather as designating what can only be designated, that is, pointed out so that we attend to it—attend as Sartre’s pour-soi attends to its object separated from it by a néant—as an object of contemplation, that is, as an exclusively mental object whose physical being for the moment is sacred and cannot be violated.
Which is, simply put, to say that the sign sets its object on a scene. But we should be guided by the Buddhist intuition that sees Sartre’s néant, the (mental) space of freedom that separates the mind from its object, as “infinite,” since the deferral of action makes my distance from the object effectively beyond my powers to traverse it. It is this equation of a limited physical distance with an infinite moral distance that offers the simplest worldly illustration of the sacred.
And in contrast, once this sacred distance has been accepted and shared by the community as a whole, its members can then come together to share the object in a collective feast, through which the prior infinity of separation is transformed into the abolition of all distance, so to speak in an infinite communal closeness. In physical terms, the change can be described as a series of physical movements; but in terms of the sacred that unites the human community, the abolition of this infinite separation is miraculous, infinitely meaningful, and effectively the source of all meaning.
In this way, Pascal’s two infinities, even more than his famous pari (bet), can be shown to embody for us with admirable concreteness his pioneering anthropological intuition of the sacred.