The Church of A.I.

I spoke at an event for Quilt.AI in The Times Center in New York City over this past week. It was a beautiful venue and an awesome gathering of people — including at least one twitter follower and substack reader who unexpectedly tracked me down to say “hi” afterwards. You make this all worth it.

I was in conversation with Gillian Tett of the Financial Times; our session was followed shortly after by an interview with the exceptional filmmaker Deepa Mehta . I learned much. But as it so often happens at these things, I didn’t have time to develop my thinking in the allotted time for our session — so I thought I’d share some of my prep notes which I didn’t have time to get to here. It starts with the wonder and beauty of ‘irrationality’, which I believe our intelligence-obsessed moment is forcing us to forget. And that has some very odd consequences…

In declaring an “educational emergency” in our lifetime, the former Pope Benedict XVI wrote that it is only through a broadened sense of reason that humanity can grapple with the most important questions of human existence.

A broadened understanding of reason is the only antidote to a reductionist sense of reason that denies itself its greatest virtues and its own teleology, its own purpose. He writes:

“A purely positivistic culture that tried to drive the question concerning God into the subjective realm, as being unscientific, would be the capitulation of reason, the renunciation of its highest possibilities, and hence a disaster for humanity, with very grave consequences.” — Benedict XVI

Reason, or rationality, is one of the main features of what I call Athens — the city of “pure reason”. But Athens wasn’t always the city of pure rationality. Socrates liked to listen to his daimon and applied his reason to questions like the definition of piety. Socrates, Plato, and their progeny would’ve looked at the rationalist community of Slate Star Codex with amusement.

Benedict’s concern was that the very notion of what reason is and what it is capable of — including its limitations and boundaries — is smaller and more circumscribed than ever before. We have boxed it in. It is often limited to science and what can be empirically verified, and it very often slides into scientism . If the pandemic wasn’t enough to show the folly of that, I don’t know what will be.

The reason of the past allowed Thomas Aquinas to probe the logos of a Trinitarian God — his reason tried to penetrate revelation, attempted to understand it better from within. To use the language of machine learning here, just for fun, he “pointed his brain” or directed his intellect at questions that fell far outside of the limited domains that we point our brains at today. He directed his intellect at questions of faith, at metaphysical questions, at intimate questions of the spiritual life.

Today’s reason is highly restricted to questions that fall only within a small subset of questions. Many fundamental questions of human existence fall outside of that subset. And they’re not even being grappled with.

Here’s one interesting example of how misguided I think the “falsifiability” trends is: many people that I’ve talked to over the past few years, including New York Times’ Ezra Klein, have gotten hung up on the idea that René Girard’s mimetic theory does not appear to be “falsifiable”.

The principle of falsifiability, of course, comes from Karl Popper, who in the middle of the twentieth century put forward the idea that for a scientific hypothesis to have credence, it must be inherently falsifiable — that is, theoretically able to be “proved” wrong.

Nassim Nicholas Taleb, who respects Popper, sees the serious limitations of Popper’s thinking. (In my view, Popper’s “falsifiability principle” is one of the most wrongly applied principles in all of philosophy.) ”Taken seriously as a theory,” Taleb writes, “Popperian falsification does not cut the mustard in an environment as complicated as ours. Popper was an improvement on Hume; but he was still writing to Mickey Mouse philosophers.” That doesn’t stop people from loving the word falsifiability. It sounds nice.

It has become a serious tell for me when people, usually with materialist worldview, start dropping the Popperian “well is it falsifiable?” line, which is like a way of trying to say “I’m an intellectually serious person” without actually saying it, or without truly understand what Popper was talking about — and the limited scope of his principle, even by Popper’s own standards.

Think, for instance, of the utter unfalsifiability of the internal states of another person. If you tell me that you’re very happy but everything you’re doing reveals to me that you’re very sad, we shouldn’t just close down any discussion about the truth of the matter because it’s not possible for me to empirically “falsify” your statement about how happy you are. It should not prevent us from exploring the truth of the matter.

Many of the most important ruths of human life are “unfalsifiable”, when you stop and think about it. Now there are certain domains in which Popper’s idea does work very well, like research into environmental and climate change. In these domains, it does have the ability to help separate real science from pseudo-science.

But dropping the “falsifiability” line at a cocktail party, or on a podcast (yes, I’ve been subjected to this) by people who want to sound smart is one of those things that tips me off to what kind of discussion it is they actually want to have. They want to put serious philosophical and theological questions on the operating table and dissect them, and then run the blood through a centrifuge until they get the lab results from Quest.

Yet serious questions like the nature of love, or the nature of the human person (fundamental anthropology), can’t be subjected to this kind of analysis. It calls for the broadened reason which Benedict saw as the key to confronting the most important problems of the 21st century.

I am reminded of the time (as recounted by Cynthia Haven in this excellent book) when René Girard was at an academic conference and approached by a person who asked skeptically how Girard could prove that was he was saying was true, since he seemed to rely so much on ancient texts. “You will see the success of my theories when you recognize yourself as a persecutor,” Girard replied.

But as long as that fundamental blindness remained — not seeing ourselves as persecutors, as victimizers — there is an epistemological barrier that prevents the truth from ever being fully revealed to us.

A broadened, expanded notion of reason is the first step in solving the three city problem — at least within Athens. That would mean, for starters perhaps, letting go of the obsession with intelligence.

I think calling what we are currently developing “artificial intelligence” is disorienting, especially considering all of the hopes and dreams that people have for it which are more spiritual in nature than intellectual.

But it’s telling, isn’t it? Calling it artificial “intelligence” gives us the illusion that it’s only our “thinking” that we’re outsourcing — and not something much more profound. It’s important that the name sticks in order for the mythology to work.

But why? Why this obsession with intelligence? The thought continues to haunt me. I believe it has something to do with the gnostic spirit sweeping through society, in which “knowing” higher and esoteric things which the “normies” don’t somehow makes you able to live a better life.

And yet intelligence is by no means a requirement to living a good life. History bears this out. Many of the simple men and women in history who lived lives of great love and died good deaths went their entire lives with the great joy of never being shared forced to share a “take” on a trivial and abstract issue.

You might say I’m a big fan of ‘irrationality’ — and that includes much of the wisdom that is embedded in rituals, the sacred, and the holy. It is ‘irrational’ only by the false standards of falsifiability. Yet with a broadened notion of reason, they begin to emerge in a logos that resists being captured in a bottle, resists being subjected to reductionistic reasoning.

Imagine if a truly tri-city AI was let loose to explore these more profound patterns in human existence without having its logic truncated, chopped off, and reduced to a materialist circularity. Until then, we are witnessing “the capitulation of reason” and “the renunciation of its highest possibilities”.

And that will indeed be a disaster for humanity with grave consequences. Reason turned back onto itself — like the examples of AI’s that write papers about themselves — is sterile, in the same way that innovation without conversion is sterile.

Give a brilliant mind constrained to Athens a billion dollars and he’ll invent the Barbie Movie, a rebranded Twitter, or pickle ball.

But put that billion dollars in the hands of the prodigal son, and he has the potential to build something truly special.