Small and Afraid and Without Knowledge
November 28th, 2022 | 10 min read
By Noah Karger
On May 11, 1997, a computer program won a chess match against a world champion for the first time in history. The defeated Garry Kasparov said that after Game 5, “he had become so dispirited that he felt the match was already over.” When reporters asked him why, he remarked: ”I’m a human being. When I see something that is well beyond my understanding, I’m afraid.”
Some 25 years later, AI is winning a lot more chess matches — and at a lot more than chess. DeepMind’s program MuZero mastered chess, shogi, and Atari games without any foreknowledge of their rules — an actual auto-didact. It even set a world record in go, which is generally considered the world’s most difficult board game, averaging 250 possible moves per turn.
And apparently, AI can create better art, too. Recently, an AI-generated piece won the annual Colorado State Fair Fine Arts Exhibition, engendering widespread controversy and confusion — everyone wondering whether it was a form of plagiarism. Plagiarized or not, the piece is breathtaking. Jason M. Allen spoke to the experience of watching Midjourney (an AI program) formulate it: “I couldn’t believe what I was seeing,” he said. “I felt like it was demonically inspired — like some otherworldly force was involved.”
In Kazuo Ishiguro’s Never Let Me Go, Kathy H. describes that moment in chess, right after a move is made, when, “you take your finger off the piece, you see the mistake you’ve made, and there’s this panic because you don’t know yet the scale of disaster you’ve left yourself open to.” In the wake of so much progress, so many decisions, something ominous looms ahead, its interpretation yet to reach consensus. Stephen Hawking, for one, wasn’t optimistic, positing that “the development of full artificial intelligence could spell the end of the human race.”
As AI surpasses us at so many of our own games, panic heightens. Lifting trembling fingers from chess pieces, one can’t help but realize — they no longer need us. We were in a rush to get here, but as experts augur an endgame on the horizon, a species-wide identity crisis ensues.
Western philosophy, theology, and society broadly have long considered the ability to reason as that which distinguishes humans from other animals. Aristotle argued that “man is the only animal whom [nature] has endowed with the gift of [logos].” Though notoriously difficult to translate, logos was often employed in ancient Greek philosophy as a kind of human capacity to think, deduce, and communicate rationally (as it is here by Aristotle).
But as recent research demonstrates, “computers are starting to reason like humans.” AI has begun modeling human intelligence in its employment of relational reasoning, the capacity to reach logical conclusions about the relationships between data, objects, events, etc. — a capacity which is fundamental to human reason. And so, we are prompted to reconsider whether logos is — in this sense, at least — distinctly human. Even if God originally gave humans alone the gift of reason, it seems we are now giving it to another. Was there anything in God’s creating us, then, that remains — and will always remain — untransferable?
Swedish philosopher Nick Bostrom argues that, because AI’s outperformance of humans is only within limited domains, it is not technically superintelligent. Still, they could become superintelligent — it is possible, even likely, but experts widely disagree on the timeline. And then there’s Rice’s theorem, which states that all non-trivial semantic properties of a programming language are undecidable. In English: “one cannot in general figure anything out about what a computer program might output just by looking at [it].” We could be staring superintelligence in the face and just not realize it — we’d have no way to be sure.
That we can no longer discern the qualities of our own technological creations signals to the changing nature of our relationship to them. Meghan O’Gieblyn notes that, whereas programmers “once saw themselves as didactic parents,” they are now more like “a holy priesthood devoted to interpreting the revelations of these runic machines,” in the same way “Talmudic scholars once labored to decipher the mysteries of scripture.” And herein lies a clue for determining the distinctly, essentially human — that fact of ourselves we cannot escape, no matter how advanced and scientific we become. The project of modernity long sought to erase it, but as we lie prostrate before the very creations it spawned, we cannot help but reassume that ancient instinct: faith.
Søren Kierkegaard once presciently said of faith that “no one goes further.” No one — not even the intelligent Georg Hegel or superintelligent MuZero — can progress past it. Unable to be understood, unable to be simulated, faith can only be lived. And as AI develops the ability to reason — or something like it — O’Gieblyn points out that faith’s unavoidability is all the clearer. Despite the illusions of the enlightenment, we can now see that rationalism — and what Yuval Noah Harari calls dataism, “requires as much faith as religious belief.”
And this is why to Kierkegaard the “essentially human” is not a rational principle, but passion. Mind you, his definition of passion is not what we tend to mean by it — strong feelings. To be sure, strong feelings are involved, but passion in its highest form is faith, and faith, properly construed, is a task — that of a lifetime.
Reason is commonly identified as the distinctly human because it is thought to be the proper faculty for seeking and telling the truth. But Kierkegaard’s conception of truth does not find ultimate expression in the epistemic certainty of reason. Rather, truth is “the daring venture of choosing the objective uncertainty with the passion of the infinite.” In this way, passion does not renounce reason, but subsumes it. What is true cannot be grasped with absolute objectivity, but it can be pursued absolutely — the adjective becomes an adverb, and that’s because, to Kierkegaard, humans are unfinished, in a constant state of becoming. What essentially defines us are processes and not their products — dispositions and not devices. Reason sans passion, according to Kierkegaard, is not truly human.
But AI is giving us an apophatic clue to what is. The essence of humanity, properly construed, is not the logos of Aristotle or the logic of MuZero; it is higher and deeper. John’s gospel opens by identifying Christ with logos, saying that “in the beginning was the Logos and the Logos was with God and the Logos was God” (1:1). Here it seems John employs Logos like the Stoics — to describe a universal principle governing the universe. But he doesn’t stop there, going on to tell us that this Logos became flesh. The universal, the transcendent, the truth of all existence — the Logos —assumed the limitations of human flesh, understanding, reason — a logos. In Christ, the universal meets the particular, the Logos intersects with a logos. And through Christ, so do we.
At this intersection emerges the essentially human. Kierkegaard depicts the human self as existing in tension, a fountain of freedom which produces anxiety and propels us to action. According to him, the human self is a synthesis — finite and infinite, possible and necessary, temporal and eternal. We are, at the most elemental level, a relation, a relation which finds ultimate expression in relating to God. Unlike MuZero, the passionate human — “the true autodidact” — is “a theodidact.” Humans discover truth, and themselves, only in passionately relating to God. But to relate to God, who is wholly other —infinitely, qualitatively distinct — we must leap, he says.
The leap of faith requires that one become the single individual before God. To do so, one must suspend “the ethical” — universal norms arrived at via reason. In this “teleological suspension of the ethical,” one does not act by virtue of reason but “by virtue of the absurd,” Kierkegaard says. The “absurd” is the paradox of faith, “that the eternal truth has come into existence in time, that God has come into existence.”
This leap cannot be performed by a machine whose existence is defined and delimited by reason, programmed to operate within the universal. MuZero works by evaluating future sequences, predicting their outcomes, and then selecting the sequence most likely to produce the most profitable result. This works within a finite set of possibilities, but in the infinite expanse of life and death, there is no most profitable result— there is only a chasm beneath us, the fear and trembling it produces, and the choice to leap.
This is a different kind of chess match, like the one in Ingmar Bergman’s The Seventh Seal — a chess match against death, life in the balance. But to Kierkegaard — and as Antonius Block discovers (who contends with death in this chess match) — life, unlike chess, “is not a problem to be solved, but a reality to be experienced.” AI differs from humanity not in the kinds of questions it asks, but in that its every action is punctuated by their predictable answers. In this way it is essentially unlike human persons, who shall not live on data alone.
Faced with the absurdity of life, the insufficiency of data to mute the loud silence of God, Antonius realizes that, more than acquiring knowledge, he longs to perform one meaningful act — a leap of faith. MuZero makes sensible decisions with sure outcomes. But what of those outrageous decisions with unsure outcomes — what of the absurd? What of life?
Saint Jude is known famously as the patron saint of lost causes. His name was seldom invoked, as people feared its resemblance to Judas Iscariot, who betrayed Jesus. Because he had so few requests, he was willing to do just about anything for anyone, even those in the most desperate of situations. Artificial intelligences don’t do anything for anyone; they do what the data tells them to. And because of this they are more efficient, more reasonable, perhaps. But picture the Godman eating bread with sailors — picture death in its grave. Is that reasonable?
That which is essentially human is not a capacity to reason, but a capacity to act beyond reason — to be present to God through faith, even in our suffering and uncertainty. This is both exemplified and effectuated in Christ, who cries out from the cross, “My God, my God, why have you forsaken me?” (Matthew 27:46, Mark 15:34). His question does not evince a lack of understanding, but rather, our’s. Quoting Psalm 22, he gives voice to our suffering. His cry is not in search of theodicies; his cry is in search of us — an invitation to be utterly present to God, even in our darkest hour. It is an invitation to be human. Christ’s cry of dereliction uncovers our real longing and its true and only answer: the presence of God. In this do we find the uniquely human, in God’s presence is our passionate existence of faith realized.
Chesterton said of atheists, “They will find only one divinity who ever uttered their isolation; only one religion in which God seemed for an instant to be an atheist.” It is the only religion where trembling fingers find their home, where the unknown is not erased but embraced in earnest. Pilot asks, what is truth? Christ responds not with philosophical deductions but death and resurrection. He responds with salvation; his answer is life eternal. This is the God of Job, whose answer is not “words without knowledge” (Job 38:2) but his own presence.
Pope Benedict XVI is right, the life of faith is “open to all that is truly rational.” Yet, the life of faith is not confined to it. Our Aristotelian logos is gathered up into the eternity of Christ, the Logos. We are thus less like the Cartesian dictum and more like Simone Weil’s: Je puis, donc je suis (I can, therefore I am). Taut between the finite and the infinite, we are driven to his presence, animated by passionate faith.
Will we be superseded? Let our answer be a prayer, like that of Christ, “deliver us from evil.” Deliver us from our devices, O God, we know not what we program. “We are small and afraid and without knowledge,” says Antonius. Yes, we are not superintelligent. We were a lost cause, but you have done the unthinkable; becoming man, you made us capacious for more than reason, for more than data. When Logos took on logos, sinners were given a higher task —a passionate one.
This task of faith can never be anything other than our own, a truth both terrifying and exhilarating. The development of AI cannot thwart it, no matter how hard we try. As we work out the implications of our artificial designs, may our fear and trembling be rightly directed, not to our inventions, but the relations we sustain, the passion we nurture.
According to Kierkegaard, the work of faith “is always adequate for a person’s lifetime,” and with it, “each generation begins all over again; the next generation advances no further than the previous one.” No matter what we outsource to machines, algorithms, and programs, Dmitri Karamazov’s words in The Brothers Karamazov will always be true: “the battlefield is the human heart.”
Mere Orthodoxy is a reader-supported publication. Support our work by subscribing to our print edition.