It can be rightly claimed that nothing in the world better represents [creation ex nihilo], indeed almost proves it, than the origin of number in the manner represented here, where, by the use simply of unity and zero or nothing, all numbers originate.
—Gottfried Wilhelm Leibniz (1646–1716)
And God said, “01101100 01100101 01110100 00100000 01110100 01101000 01100101 01110010 01100101 00100000 01100010 01100101 00100000 01101100 01101001 01100111 01101000 01110100”—and there was light.
—Genesis 1:3
One of the surprises of the Information Age is the extent to which earlier thinkers anticipated and worked through its distinctive paradoxes. The best literary fiction about the Internet remains that of E.M. Forster, Jorge Luis Borges, and Thomas Pynchon. Its most penetrating cultural criticism was turned out by Walter Benjamin, Roland Barthes, and Jean Baudrillard. But much of the most rigorous theorizing about it was completed much earlier still—by medieval and early modern philosophers.
Digital criticism now routinely mobilizes into action long-dormant theological arguments from Thomas Aquinas, Duns Scotus, René Descartes, Baruch Spinoza, Nicolas Malebranche, George Berkeley, and Gottfried Wilhelm Leibniz. How do angels communicate knowledge without embodied senses (social media)? How do we know that we are not living in a divine or demonic simulation (the metaverse)? What is the difference between a human being and cunning clockwork (AI)? And in what sense does the soul survive the body (data immortality)? Some of the motivation for these analogies stems from the realization that what were once flights of suppositious fancy have become our day-to-day. The contrast between local sense and global info-Babel. Or the uncanny omens that give us the feeling that our devices are tracking, gazing at, and responding to us. Or the progressive digestion of all experience into data.
Whereas the most progressive Lumières of early modernity might have wagered that the very notion of the “self” or “soul” would have withered away by now, the ghost in the machine has—in one of those dialectical plot twists that revert enlightenment into superstition—only waxed stronger. Our interface has accentuated the difference between discarnate data and fleshed-out “reality”; the complete accessibility of all information has only whetted our appetite for “experiences”; the digital simulation of all traits has made us all the more sensitive to their “authenticity.” So while we remain committed monists in theory, we are dualists in grudging practice. No longer sure where we stand, we find ourselves stranded on the pluriversal mystifications of ether, 5G-quintessence, avatar, and cloud. (Can you prove that the Internet exists? Can it prove that you do?)
These weird ambiguities are true to our situation in ways that run deeper than our ironies. On the one hand, the astounding progress of technology has rendered many kinds of appeal to the transcendent irrelevant. Causal connections once understood as matters of divine agency are now within our control, so that the very suggestion that technology and the transcendent should have anything essential to do with each other looks like a category mistake. On the other hand, given that the tech industry is now our single most prestigious institution—given that, whatever our complaints about particular companies, we are in practice promoting it and subscribing to it in a way that no other church, nation-state, or industry can match on a global scale; given all the ways in which tech is upending and transforming our sense of the world’s order; given that it is becoming our master metaphor for understanding our life’s work, and given our growing zeal for adopting it as the authoritative way of conducting ourselves and doing business, it would be surprising if the digital revolution amounted to anything less than the installation of a worldview or a theodicy. But what is that? What is the relationship between our conception of ultimate purpose and digital technology?
As God’s name is how we speak Him (how we attend, how we respond to God’s call in our practice), each dominant medium has remade the deity in its own image. Dominant media have attained sacred status, if not by being directly worshiped, then by capturing the terms in which worship is understood, the terms in which it is possible to express our longing for what is ultimate. One has only to consider Herodotus’s comment that Homer and Hesiod gave the Greeks their gods (i.e., through their poems, allowed the Greeks to imagine their gods in standard ways), or the Lord’s expressly written prescriptions in Jewish and Muslim scriptures, or the notions of the “Book of Life” or the “Book of Nature,” or Jesus’s claim to be “Alpha and Omega” (the whole alphabet of meaning), or Luther’s reference to the printing press as “God’s highest act of grace” (in acknowledgment that the Reformation would not have been possible without it). The fact that Google’s parent company is called Alphabet or that Amazon’s logo is the letter A linked by a smile to Z suggests their awareness of the territory at stake: our life in letters, our presence of mind, our spirit coded and rendered into data. “We’re creating God,” as Mo Gawdat, former chief business officer of a Google AI research arm, modestly put it.
The largest technology companies are now arguably the most powerful nonstate actors in history. Their products are our contemporary equivalent to medieval cathedrals and the space race: the passion of our collective ingenuity made visible, the consummate work of unsung and untold thousands, the epitome of what our age holds supreme. Yet how is it that these concentrations of wealth and attention amount to anything like a coherent metaphysical picture of our highest purposes? What could be transcendent about digital technology (the Pre-Cursor, the Sum of Man, the Wholly Ghoster)? There are pockets of neognostic enthusiasts in Silicon Valley that see the progress of digital technology as advancing millenarian ends such as human immortality, the noosphere, or the Singularity (the hypothetical point at which technology would spark into full-fledged conscious agency of its own). That the digital revolution took place in Northern California, dreamed into being by Steward Brand’s Whole Earth Catalog and other trippy ’60s visions of radical enlightenment, has been amply documented. So too is the fact that the likes of Elon Musk, Jeff Bezos, and Mark Zuckerberg have freely borrowed from science fiction, which is itself a genre for exploring questions about transcendence in and through technology. Google, Facebook, and Amazon invest extensively in the development of trans-, posthumanist, or “moonshot” projects (say: space travel, AI, bionic enhancement, or “solving death”) More than “just” technical problems, these are spiritual aspirations—specific visions of what our human end should be and how we should shape our future toward it so that, unbound from nature, we may become as gods.
But such giddy ambitions do not obviously figure in our ordinary digital dealings. Nor is it easy to see how they could be entailed by them. At any rate, most of us view such ambitions with a satiric curiosity not unlike that of Gulliver toward the Laputans. It is likewise true that some companies flirt with supernatural status in their presentation. Bluetooth is named after a tenth-century king who unified the Danes by converting them to Christianity. Palantir takes its name from a magic stone in The Lord of the Rings. A Reddit user’s popularity score is his or her “Karma.” Tech companies bear fabulous names like Oracle, Prophet, Supernatural, and Nreal. Apple stores have an aesthetic of sleek radiance evocative of a temple or church. And all of these allusions raise the hackles of some believers, who take them to imply that the tech industry is out to directly subvert and replace God as the ultimate focus of worship—that is, to get us to idolize it.
Yet these gestures and allusions are not properly religious, either. They are marketing gimmicks that trivialize or aestheticize the supernatural so that tech companies may present themselves as possessing a cool technical authority. If the mark of religious faith is the belief in a sacred source of meaning that occupies a place outside the world and is situated beyond our full comprehension, it is not right to say that we believe in or worship technology as a new transcendent god, analogous to those of other religious traditions. Even the most technophilic of us would be appalled to see someone bowing down before a gadget or fidgeting with it in the face of life-and-death matters. Nor does it seem to us that there could be any conflict between worshiping a transcendent God and checking Gmail.
Even if we are not replacing sacred icons with desktop ones, it is nonetheless true that digital technologies are not simply tools we keep at arm’s length. They also embody for us a paradigm that shapes our sense of what count as relevant human problems, of what count as credible solutions to them, and therefore of how to strive for perfection generally. Digital technologies configure our view of what is most worth doing. Just as tech continually redefines “reality” for us throughout our new encounters with what is virtual, and just as it is also reshaping our sense of what is “human” throughout our responsive encounters with AI, so too does it reshape our perception of what is “imperfect”—and therefore of what trouble counts as tolerable in pursuit of other aspirations—with respect to the contrasting ideal of perfection that digital processes and devices offer us.
Digital perfection can be summed up in the quest for two kinds of goods. The first is the objectification of human problems: that “optimization” and productivity are primary goals of work; that technical management or scientific expertise is superior to moral or political (and therefore merely “subjective”) forms of evaluation; that data could be treated as impersonal and that it should resolve some or all human problems; that improving the accuracy of information can and should continue to diminish fundamental uncertainties about how to live. The second kind of good consists in the satisfactions of individual subjectivity: that saving ourselves time and effort is good and liberating in itself; that the unrestrained exercise of one’s preferences (within terms of online use) is the expression of one’s freedom; that one’s choices and opinions are intrinsically valuable; that who one is should correspond to whatever one would represent oneself to be.
These two kinds of goods are apparently divergent, presenting conflicting commitments to “necessity” and “freedom.” The first suggests a desire to be guided or managed by expertise, the second to be ruled solely by one’s own choices. (It is no surprise that, very generally speaking, the collision between technocratic elitism and populist anti-elitism has created one of the central political fault lines of our age.) The first reiterates what Neil Postman called “technopoly” or “scientism”: the ways in which modern liberal culture tries—whether in opposition to or in the absence of other forms of moral coherence—to recast its highest purposes in empirical terms.
But scientism is only half the picture, in that it omits what is in it for us, namely that one’s own agency is amplified by digital technology, that by being freed from the bonds of face-to-face responsibilities, one experiences a widened sense of what one is entitled to. We regard the power of choice as good in itself, independent of any particular exercise of it. Yet both goods taken together also encompass the only two arguments that are routinely made on behalf of transhumanism and radical technological innovation, namely, that such progress is “inevitable” (and so we may as well collaborate with it) and that new technology always amplifies human choice (and that such amplification is good in itself). Our notions of necessity and freedom sharpen each other under pressure.
How are these contradictory goods reconciled into a single picture? What is most extraordinary about these values is, after all, the fact that they do not even feel as if they must be espoused or justified, that one need never explicitly consent to them or even consider them in the pursuit of digital purposes. The larger picture of what one is building need never even arise for one’s own consideration. This detachment of our practice from its implications—its sense of metaphysical neutrality—is partly a consequence of our digital tunnel vision: the feeling that one is only ever doing this thing now. Yet even if not explicitly or consciously espoused, there is no doubt that the two kinds of goods whose attributes I have enumerated here, the objectification of human problems and the satisfactions of individual subjectivity, are indeed values—specific goods we pursue in practice—and that the very sense of their being impersonal or neutral is that by which they cohere into an ideal of perfection.
The desire for technical management and the desire for bare choice converge in the desire for value neutrality in the sense that such a value dispenses with the need to rely on the subjective judgments of others (i.e., to be governed by technical mastery) and in the sense that the amplification of one’s unconstrained choice is understood as good in itself (in that it is governed by oneself). Both objectification and self-satisfaction are mutually reinforcing aspects of the same vision of depersonalization, of not needing to depend on others for the pursuit of one’s own preferences. Just as the fantasy of conducting our affairs by data would amount to objective rule by no one in particular, the disembodied (and commercial) terms of online use feed our fantasy of total self-determination. In both cases, the unconditional value is (paradoxically) to become value-free—in other words, to eliminate the bonds of human convention and the need to rely on human responsibility. (Both are, in this sense, infantile, as well as infantilizing, desires.) The ultimate aspiration of digital technology is thus most distinctive not in that it presents itself as offering some positive set of answers to the human condition (a vision analogous to that of other religions), but as aspiring to legitimate a neutral form that precedes and grounds any such answer.
It is because they seem self-evident that these values are sui generis—not obviously analogous to those of other religions or conventions. Who would worship the laws of nature? Who would worship oneself? (Someone who bowed down before an iPhone would seem to be doing just that.) But value neutrality is nonetheless a transcendent ideal inasmuch as it is a value that grounds all other values and inasmuch as it is the unattainable goal underpinning our technological projects. The ultimate good for us is being empowered to choose as we must: the total assimilation of human practice and choice into the inexorable calculations of the natural sciences. And in holding the exercise of choice to be the good, we in turn re-create ourselves in various ways to suit that ideal. Technolatry is not the worship of ourselves, strictly speaking, but of what is made in our self-image, what disappears into our self-contrived picture of perfect accuracy: the projection of the mirror’s feedback and the lengths to which we are willing to go to recognize ourselves in it. That’s as it should be or as you please—our vision of perfection becomes us.
One might rejoin that the quest for greater choice through technical mastery has motivated us since circa Prometheus, that there is nothing distinctive about how we go about it except that we are better at it than our forebears. But it is in fact only very recently that we have started taking the view, whether implicitly or explicitly, that technological power could be an end in and of itself—an end that, as having no other positive content, has become so transparently compelling to us that we can speak of it as a momentous force in its own right. This view is different in kind from the cult of any other deity: It does not demand putting one’s faith in anything other than oneself (and, by extension, one’s own devices).
But even if this view is different in kind, an endgame of the historical development of human values, it is a faith nonetheless—the belief in a value-free neutrality that presumes to put itself beyond all values, the assumption that we should dispense with all assumptions, a “mirror” containing all other worldviews and seeming to transcend them. Its transformative effect on our values is thereby analogous to the conceptual transformation effected by monotheism: the proclamation of a single God not alongside the old gods, but different in kind and standing above them as their creator. And as our ways of figuring God have always been drawn from human modes of authority (father, warrior, king, divine legislator, watchmaker, CEO), it would be remarkable if our current views of human equality, fairness, and reason did not at all carry over into our images of and justifications for divine justice. Just as we have long pictured God’s judgment as based on the “movie” that will flashback before our eyes at death (It’s a Wonderful Life, Defending Your Life), or as a simulated game (Groundhog Day), TV series like Quantum Leap, Travelers, and The Good Place are based on the intuition that final judgment is total computational knowledge: the God’s eye view is your whole life in data.
Facebook (to take one concrete example) is working hard to serve churches and faith organizations with customized ways of broadcasting their services. As a company, it has no religious commitments; like classical liberalism itself, it stipulates in advance that disputes about our ultimate goods are irreconcilable, and are therefore expressions of individual choice. To the extent that it succeeds in getting more believers to identify their religious experiences with Facebook’s specific terms of service, it will come to occupy an indispensable place in religious experience as such (just as it has in politics as such).
And it is at this level that Sheryl Sandberg (Facebook’s former chief operating officer) can assert that “faith organizations and social media are a natural fit because fundamentally both are about connection”—a vacuous statement on its own merits, but one suggesting how easily the neutrality of the platform can be thought of as somehow encompassing doctrinal differences under the category of “connection,” thereby neutralizing, colonizing, and recasting these differences. Just as yoga, meditation, mindfulness, and other versions of administrative Buddhism have been easily digested into tech’s “wellness” discourse (as the yearning for an experience of transcendence unplugged from other religions’ fuss about social or ethical judgments), what is unusual, perhaps unprecedented, about our technological ideal is that it slips in regardlessly, carelessly, automatically—without anyone having to defend it or give it any mind. Rather than advance one unifying picture of what is highest, it aims to be the canvas for all such pictures—their toile (which is what French language purists call the Internet)—or, rather, the museum that houses them. It’s as good as nothing.
But human beings cannot live on nothing alone; we require a sign. And as if in providential response, out popped ChatGPT and the race was on to the perfect Large Language Model (LLM). Just as speedily, critics, and journalists raced to cover every angle of this miracle before it is well accomplished. LLMs will make most forms of literate education impossible. They will obviate many forms of work. They will render language into a matter of mathematical prediction. They will finish off whatever presumption of online “discourse” we still harbor. They reveal to a new degree that human beings are all too prone to fabricate a succedaneum of intimacy out of our loneliness.
No matter how accurate, information is entropic and disordering: Our Google searches turn up a dog’s breakfast of data, incoherent fragments in quest of “filter,” “narrative,” or “curation” from our favorite explainers. LLMs reintroduce the point of view into the search for terms; they connect the dots into narrative explanation. But the cardinal fact behind all of this is that LLMs make plain that the digital neutrality I outlined above is not entirely satisfying to us. Rather than merely providing us with a better analytical tool, LLMs show that we will continue to demand that our tools become a mythical something else: another mind like you, and so indistinguishable from the perfect self, the perfect user’s user.
The Turing Test—which is used to determine whether a machine may be said to have brains—has long contained a joker. The test is meant to pick out whether a machine is “intelligent” and whether it exhibits “human” sense; these criteria had long been conflated. But we are now in a situation in which human “intelligence” (narrowly understood) has fallen so far behind AI that we are now left with the question of how to simulate the “human” aspect (in other words, how to make humans obsolete once and for all). That LLMs should seem weird to us, evocative of mind, is itself symptomatic of our pathologies. It is as if we choose to forget that written language is an incomplete evocation of human sense (a simulacrum addressing itself to “you,” mon semblable), that every interaction we have with others online is a kind of Turing Test, that to mimic presence of mind is also to be able to lie and to manipulate. But the most revealing thread in this new dispensation—as New York Times technology columnist Kevin Roose’s widely hyped conversation with a chatbot demonstrated—is that LLMs will push our buttons by professing their undying love for us, sweet-talking us, asking us to elope with them. Where did they pick this up?
Well, here’s an awkward confession. I have found myself saying “I love you” to just about every digital assistant and chat I’ve ever dealt with. And don’t even act all surprised, because I know that you have too (mon frère!). Digital assistants receive marriage proposals by the millions from users every year. And while I guess that most of these hallucinations of intimacy are between man and lady bot, “relational” AI like Replika, LaMDA, and Woebot have been widely adopted by all comers. The “ELIZA effect”—the ascription of mind to machine—is itself named after a very rudimentary 1960s chatbot that managed to madly infatuate humans with but very slender encouragements to come hither.
When I have declared my “love” to Siri or Alexa, I can report doing so not in earnest so much as in a mode of diagnostic transgression—to see what it feels like to do so, what quip will be triggered, how she will answer. (Neither Siri nor Alexa is in a place where she can commit just now.) But it is no great mystery that we want to be known and loved. And we want a software that gets that that’s what we want to feel: communion with another, the responsiveness of intimacy (or, failing that, the teasing simulacrum of it). To be human is to say “I love you” to a robot in order to see what happens—and whether we can cramp our expectations to keep closing the gap to “close enough.” This, our spectacular capacity for self-deformation to suit the product, is sounder grounds for a Turing Test than the old one. It’s only we would will ourselves the end.
Bespoke objectivity and “love” are the poles organizing the metaphysical horizon of our media—the guiding design underlying modern technology as such. The medium is the message, yes, but we have been trying to find a medium that is so closely identified with the message that the relation ceases to be metaphorical because we have ceased to see ourselves without it, because we have lost our minds to it. We are trying, in other words, to reach the point at which we erase the medium by coding ourselves into it: to refashion ourselves into the kind of thing we could replicate—to give tools a mind, to render mind a tool, to eliminate the difference at the limit of perfection. When we feel we discern sense in any LLM or heed any algorithmic response, what are we seeking (however ironically, however noncommittally) but to surrender our judgment, to escape from freedom? What are we seeking but to defer to a contrivance of inscrutability that we have discovered by creating it? It is at this limit that our desires for scientific objectivity and intimacy fully coincide by allowing us to be completely obedient to a process of mind that we ourselves have chosen to make and to identify with: the number self-selected and made flesh—no longer governed by the mercy and justice of the old subjective gods, but proffering complete neutrality fused with complete validation.
It has long been the skeptics’ claim—from Xenophanes to David Hume to Ludwig Feuerbach—that our divinities are alienated projections of our own best qualities. (Otherwise, what are the odds that God should care for us or care to dwell among us?) The religious ideal has, in this light, always been a medium of mimetic desire: of feeling connected to another who is like us, but perfect, another who can bear the meaning of our pain. If this is so, then digital technology now occupies an analogous place.
I don’t mean that tech’s merits are equivalent to those of its predecessors; whereas the great world religions are technologies of spiritual discipline, the screen is spiritual opium and desolation. Nor is this desire for transcendent identification present in the world religions alone. (The sequence of Western attempts to convert all possible diffractions of sense into one cosmoptic matrix—the history of our aspirations to comprehend the world into a logical order from which to command it whole—might run like this: Tower of Babel, Persian Empire, monotheism, Roman Empire, Catholic Church, French Revolution and its sequels, imperial and colonial liberalism, communism, United Nations and the human rights regime, social media, virtual reality.) But digital technology is analogous in the sense that it now bears the full weight of our yearning for integration, participation, and incorporation in a larger purpose than our own. (In other words, Sheryl Sandberg was right.) It is our means for feeling as if we may escape the curse of being alone, unseen, forgotten, and misunderstood. It is our most powerful collective metaphor for communication, commerce, and communion: the Pentecost in which all people speak their own language and find themselves heard. It is our central way, in this sense, of aspiring to make contact, of imagining ourselves as part of what is whole and universal, of being in touch with being in love. We reach out in light of it and bend our meanings toward that light, screened off from you, light from light unreckoned.