Technological products 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.”
Antiqua et Nova, §41
Setting aside for another day a wide range of general issues associated with AI use (intellectual property theft, environmental damage, inequality of access, the prospect of mass unemployment through economic disruption and more), we can see from the passage above a particularly serious reason why LLMs are faulty vessels specifically for prayer, meditation, spiritual counsel or reflection. It is no superstition to suspect that they would be likely to convey a bad spirit; rather, it is but to recognize the rational consequences of the worldview from which they spring.
In February, the International Theological Commission published a document Quo vadis, humanitas? (Humanity, where are you going?). This work addressed “the recent acceleration of technological development and scientific progress.” AI was prominent among the issues addressed, with special cognizance given to AGI (Artificial General Intelligence, long referred to as “strong AI”).
For any reader unfamiliar with AGI, this hypothetical future AI is expected to match, or rather exceed, human capacities en bloc. As “strong AI,” it was represented as the achievement of consciousness and self-awareness in AI; the move away from that term does not indicate a new modesty in objectives, but rather a pervasive skepticism about the meaningfulness and value of consciousness as a goal. §38 of Quo vadis, humanitas? defines and describes the problematic:
AGI refers to a future, pervasive technology capable of replacing all computational and operational aspects of human intelligence thanks to extremely high computing speeds, made possible by the future development of quantum processors. Where specific aspects of human intelligence are consciously weakened or abandoned, AGI could have profound consequences that risk escaping human control. Some therefore hypothesize that AI will be asked to manage the problems it has created, according to a dynamic that would become irreversible. This vicious circle is presented by some as an almost necessary process, as in natural processes, a destiny that would bring to extreme consequences what human beings would themselves have originally caused.
Clearly, one strain of the worldview behind the development of this technology is Prometheanism. However, “the focus of [Quo vadis, humanitas? is] an engagement with the challenges of trans– and post-humanism.” These issues, which have concerned us for some time, are brought to a head in §41 of the document:
There is an intense race to increasingly bridge the gap between the self-improvement of AI systems and the achievement of AGI. Although the goal is still far off, it is being pursued with great determination, but sometimes without the caution generated by the wise recognition that good always involves appropriate limits and proportions. Whether it is a distant dream or an imminent innovation, this form of general AI stimulates the search for a deeper understanding of the nature of human intelligence, its uniqueness among living beings, its irreplaceability, especially in relation to moral responsibility and its intrinsic openness to transcendence. A type of knowledge and calculation that dispenses with intelligence that is experienced in a body and situated, as well as with a type of relational knowledge which is transmitted from generation to generation through educational processes that play on identity and the meaning to be given to one’s destiny and role in the world, constitutes a threat to the true good of humanity. Yet, the dreams of transhumanism, which in posthumanism even imagines an evolutionary leap, are based on this type of knowledge without a body, without limits, without ties, and without moral sense. Such imagination forcefully raises the question of the ultimate goal of technological progress.
The whole package unravels, terrifyingly: knowledge and calculation radically individualized, arelational, “without a body, without limits, without ties, and without moral sense.” In this vision, knowledge, calculation, and (supposed) intelligence held as values in themselves are ordained ultimately to transform beyond recognizability (transhumanism) or even supersede (posthumanism) human life and human existence.
This, of course, is far off, highly speculative, and perhaps fantastical – light years from any ChatGPT-generated spiritual blog post or online article. Yet there are true believers among the AI elite, for whom today’s AI is but a steppingstone to this putative future, and their worldview trickles down to the work product on the screen of the everyday end user. Even the fact that AI text is designed not around the truth that is found in being, in things (which is thus inaccessible to an LLM), but around a linguistic weave of persuasiveness, plausibility and sycophancy – yes, even that is morally toxic: could anyone imagine taking spiritual counsel from a priest or religious whose character was dominated by these traits?
The Church offers a far different wisdom. Antiqua et Nova states that “human intelligence is a diverse, multifaceted, and complex reality: individual and social, rational and affective, conceptual and symbolic.”
The Christian cosmos is sacramental; it rests in signs and types. The inherent spirituality of all things manifests itself on the razor edge of perceptibility, precisely inasmuch as being is not univocal. The work of the intellect is to divest what has been received through the senses of its individuating materiality, and know in it its core universal. But there is more: the passive intellect now moves within this immaterial architecture, “becoming” that which it has beheld. Thus, there is an immaterial, but real and experimental, contact and identification with objective reality, that amounts to incomparably more than a mere reservoir of data points. This passive intellect (together with the affectivity) is the very locus of contemplation.
To see winter sun catching on ilex, hear music in the Dorian mode, taste an Arctic Joy peach off the tree in August, is to know something, something infinitely greater than the specific, individual realities that have conveyed this knowledge. Entering into oneself to transpose interior knowledge to words and syllogisms is painful and often solitary work – yet without this transposition, we have only ephemeral impressions, but no replicable, transferable wisdom – a thoroughly human technology, if you will.
According to Antiqua et Nova, “True intelligence is shaped by divine love, which ‘is poured forth in our hearts by the Holy Spirit’” (Rom. 5:5). From this, it follows that human intelligence possesses an essential contemplative dimension, an unselfish openness to the True, the Good, and the Beautiful, beyond any utilitarian purpose.” Thomas, in deciphering what it means to say that the angels cry, cites their intensity of desire; words express the depths of a person, his ardor, his vehemence. This type of knowing and its concomitant expression are proscribed to any device; whatever may be generated to conceal this (often through the kind of cringy and contrived breathless syntax so brilliantly parodied by a friend of our Substack), LLMs are permanently rooted in a flat neutrality.
Finally, it is not surprising that the machinations of those who do not acknowledge the primacy of our lives as psychophysical beings with free and rational minds should result in the wholesale cheapening of human thought. Words are sacred. They are true to the degree that they conform to the inner word held by the soul. Intellectual mastery happens only insofar as one wrestles, like Jacob with the angel, with bringing expression to birth against the ex nihilo of an empty page.[1] Novelist and MIT professor Micah Nathan expressed this dynamic with insight and precision:
Writing… isn’t supposed to be easy, and of course it can be tedious but that doesn’t make it rote. Writing isn’t just the production of sentences – it’s the training of endurance by way of sustained attention. It’s a way of learning what one thinks by attempting to say it. An LLM can reproduce the appearance of that activity, but it can’t replace it, because the value lies not only in the object produced but in the transformation that occurs during its making.
Skipping that process is singing with auto tune, all polish and no substance – the anti-image of the eucharistic, which is wholly and ever substantial. AI, based as it is on utilitarian materialism, at once deforms the would-be writer and provides the means of his corruption ready to hand.
In this new day, anyone with an idea and a prompt can outsource a framing that would seem to indicate education and ratiocination at a depth level. Any idea, however holy, however invested with nobility and potential, is valueless unless it is actualized through the very great discursive labor of articulating and enshrining it in language. To generate a “spiritual” reflection through an LLM can do no more to build up the soul than a Zoom background of a Left Bank café setting can do to acquaint one with Paris or nourish one with café au lait and a warm croissant. Even aside from the difference in spirit between the AI mindset and the mind of Christ and of the Church, the issue is simple: you cannot give what you do not yourself have. So, to those who share our faith in Christ, we say: feed yourself and others with the daily bread of your textual and spiritual labors, not the etched electronic silicon stones of ChatGPT religiosity.
Image: Computer chip, Wikimedia Commons
[1] An aside: I know from direct experience that reading and attending lectures on metaphysics (as scintillating as they were!) did not cause me to learn metaphysics; they caused me to think I learned metaphysics. The pressure of a blue book on the desk during finals, needing to capture the distinction between act and potency, say, is where the rubber hit the road. The heavy lifting, part quietude and part deep conversation, all recourse to the sacraments and the masters, is the basis on which I have something to say and can say it. -VJT
Dr. Paul Chu is currently a philosophy instructor for CTState, the Connecticut Community College, and has previously taught philosophy in college, university, and seminary settings. He also served as a staff writer and editor for various national publications. He is co-founder of Sacred Beauty, a Private Association of the Faithful in the Diocese of Bridgeport dedicated to honoring the beauty and holiness of God through artistic and intellectual creativity founded in prayer, especially Eucharistic contemplation. He contributes regularly to https://questionsdisputedandotherwise.substack.com/.
V. J. Tarantino is co-founder of Sacred Beauty, a Private Association of the Faithful in the Diocese of Bridgeport. She has studied ancient and Medieval metaphysics and has devoted her adult life to the service of liturgy (study of liturgical texts and norms, the cultivation of sacred elocution, musical performance and composition, the beautification of sacred space, and the organization and direction of public Eucharistic Adoration) and to immersion in the writings of the Doctors of the Church and of recent Popes. Her writing can be found at https://questionsdisputedandotherwise.substack.com/
Popular Posts