Recently, in response to some of my own writing on the topic, Andrew Caspersen, a reader of the Abbey, sent me an article he had written on how Christians should respond to AI. This is a topic I struggle with all the time, and part of the struggle is how to articulate the challenge in Christian terms. I thought Andrew’s piece did this so well that I invited him to publish it here. I would love to hear the thoughts and responses of other readers. Please do leave your comments below.

Paul

I am a Catholic, father and business contractor living in New York City. I’m also an ex-convict, rescued by our Lord’s divine mercy. I wrote this piece because nearly every anti-AI argument I’ve read concerns its risks. This concerns its very nature.

- Andrew Caspersen

Long before the digital age, artists have pondered both the promise and peril of artificial intelligence. Homer’s Iliad described automaton tripods made by Hephaistos that might be confused with Elon Musk’s dream of putting an Optimus robot in every home. Ovid’s Metamorphoses tells of Pygmalion, disappointed in his marital prospects, carving his perfect woman and bringing her to life with the help of Venus. She was the first romantic chatbot, a reflection of Pygmalion’s self-love. Mary Shelley’s Frankenstein warns us what can happen when scientific hubris puts the creation of life and intelligence in its crosshairs. More recently, we have Frank Herbert’s Dune, with the maxim “thou shall not make a machine in the likeness of a human mind.”

Yet we are here, crafting such a machine. Many, perhaps the majority of us, have deep reservations about AI, yet struggle to explain why. We think it could create trouble but also see the great benefits. As a result, nearly all the arguments have been around the outcomes of AI. This is the modern approach - judging something by what it can do, rather than what it is.

If artificial intelligence were a tree, then nearly every discussion we are currently seeing concerns its fruits. An AI optimist predicts a cure for cancer, where a pessimist sees the elimination of the doctor-patient relationship. Mass prosperity! Or is it mass layoffs? For every rotten apple I spot, another sees a juicy golden delicious. How can we make sense of it all?

Arguing about the fruits is problematic. First, predictions are tricky business. Remember when social media was supposed to bring people together? Second, Christianity focuses on the means, not the ends. We care more about whether eggs are getting cracked than how tasty the future omelette may be.

Let’s go from the fruits to the branches. Are there certain features or applications of AI that are inherently wrong? Yes! Marc Barnes uses natural law to show why conversing with a chatbot is a corruption of inter-personal communion. D.C. Schindler explains the essential duplicity of AI, which tricks the subconscious into accepting its personhood.1 These arguments reveal the rot in some of the branches of the AI tree.

They got me thinking: is there an argument against the tree itself, from the roots to the fruits? Perhaps it goes something like this.

Imago Dei

God created man in His divine image with the gifts of knowledge and love, according to the Catechism and saints from Augustine to Aquinas. The imago dei is the realm of the intellect and will. The mind and heart.

The mind is capable of many things, including reason, wisdom, understanding, conscience, memory, imagination and judgment. The mind calls the will to action. It is the finite image of the infinite Word.

What should I do with this awesome gift of the mind? I really have three options. Pursue God’s will, oppose God’s will, or not use my mind at all, due to distractions. Much of my life has involved choosing doors two and three, with the predictable consequences.

Artificial Intelligence now offers me a fourth option: an enhanced mind. AI does this through simulating ideation, composition, artistic creation, analysis, recommendation and conversation, to name a few examples. And since these simulations are often better and quicker than the work of my own mind, it is very appealing to outsource my thinking. The fact that it comes with the carrot of career advancement and the stick of unemployment makes it almost irresistible.

But there is a price. By accepting this man-made mind, I would be treating the divine gift as inadequate. And because the divine gift is an image of the Giver, I would then be treating God as inadequate.

Part of me wishes I did not know that. Then I could simply use AI, not knowing what I was rejecting. Yet, like the false security offered by Dostoyevsky’s Grand Inquisitor, it would be costly ignorance.

God is all-powerful, all-present, all-knowing and all-loving. Mercifully, He didn’t share his power, for power without wisdom does not end well (another argument against AI). Instead, he endowed us with knowledge and love. Why? So that we can come to know Him and love Him. And in so doing, we are drawn to serve Him in this life and dwell with Him in eternity. That is the meaning of the Christian life. Our telos.

What I’m presenting is a very simple Christian argument against AI:

- The human intellect was made in the divine image

- AI seeks to replace at least part of our intellect in practice

- Therefore, using AI would be a rejection of part of our divine image

- A partial rejection of the divine image is a partial rejection of God Himself, one small act at a time

Potential Counter-Arguments & Rebuttals

Counter-Argument #1

AI isn’t real intelligence, it’s artificial. The artificial can never truly replace the real. By accepting AI, we are not replacing our own intellect. Therefore, we are not rejecting part of our divine image.

Sed contra

It’s true that AI isn’t real (the adjective ‘artificial’ is a dead giveaway). It’s also true that the artificial can never truly replace the real in essence. But it can still act as a replacement in practice. Artificial ultra-processed food may not be a true replacement of real food. And yet, it ends up replacing it in practice. Pornography is not a true replacement of real love. But now we have millions of young men viewing pornography instead of engaging in true romantic relationships. The artificial always competes with the real. That’s how it’s designed.

Counter-Argument #2

AI is not a rejection of our intellect at all, but instead a complement. AI can help me think better, process more information, be more productive for my family, etc. I am still using my intellect just as much, if not more. I am just getting help with some of the blocking and tackling. I’m still the quarterback.

Sed Contra

Every time AI is used to ideate, reason, write, create, analyse, recommend, etc., it is performing the mind’s work. In that particular instance, it is not a complement but a substitute. And the more we use it, the more we will be replacing the divine gift, one small act at a time.

Counter-Argument #3

AI is just the latest in a long line of tech that makes our mind’s work easier. Sure, it’s more powerful than the abacus or Excel spreadsheets or word processors. But the intention is the same - to make our work more efficient through the use of tools.

Sed Contra

Yes, we have used plenty of tools to assist in our mind’s work. But this is the first tool that is actually designed to mimic our mind itself. It is not a difference in degree but a difference in kind.

Counter-Argument #4

By using AI to do the small things, I can free up my mind for the really important things. I will even have more time to spend time with loved ones, to volunteer, to pray and to go to church.

Sed Contra

This seems both unlikely and incidental. Unlikely because all the evidence points to digital technology crowding out these kinds of family and spiritual activities. Incidental because it does not deal with the core issue - whether using AI is a (partial) rejection of our divine gift. If it is such a rejection, then it is wrong, no matter the benefits one person receives.

Counter-Argument #5

AI is inevitable. Resisting won’t make a difference. It will be everywhere and you will be disadvantaged. If your kids don’t use it, they will be disadvantaged.

Sed Contra

Yes, this seems inevitable at the societal level, regardless of any one person’s resistance. Recently, my boss asked me why I wouldn’t use AI. Once I explained my beliefs, her response was incredibly charitable. She simply asked, “aren’t you afraid of being left behind?” If I’m honest, I would say that I am a little afraid. But being willing to be left behind by the world has always been part of the Christian bargain - forsaking the goods of this world for the Summum Bonum.

Counter-Argument #6

Christians have always used the latest technology to preach the Gospel. Like the Roman roads, codex, and printing press of centuries past, AI can be used to engage the next generation of converts.

Sed Contra

This argument brings us back to the ends, not the means. I have no doubt that AI could yield many evangelical benefits. Catholic AI companies like Truthly promise “instant answers” from “2,000 years of wisdom.” This approach promotes knowledge as a collection of facts. The saints meditated on the Word day and night. Today we can simply punch our questions into a computer. I’ll cast my lot with the saints.

Counter-Argument #7

AI will never be conscious. It is just a sophisticated computer. How could that possibly replace the imago dei?

Sed Contra

I certainly hope it won’t be conscious. If it were, it would mean (i) we’ve ushered in something and (ii) we’ve attempted to enslave that something through alignment. That is a terrifying prospect which is outside the scope of this argument.2 But assuming AI will never be conscious does not free us from the imago dei problem. We would still be using something that is built in the image of the mind to do the mind’s work. We would be substituting the imago dei with the imago hominis.

Putting on the Mind of Christ

If I’m wrong about AI, then at least I’m doing it for love of God. I’ve been wrong my whole life, so why stop now? But if I’m right, then AI is not a new story at all but one of the oldest. God creates man in His image. Man is not satisfied. Man creates god in his image. Trouble ensues.

What kind of trouble? This returns us to the realm of predicting fruits. This is tricky business, especially for an ex-gambler like me who never bet he would end up in prison. To be safe, I’ll focus on the one thing that truly matters. How could AI impact my relationship with God?

God gave me an intellectual soul to contemplate His love, meditate on His mysteries, speak to Him, adore Him, beg for His mercy, understand His will and ask for the courage to pursue it. If I set aside this divine gift in favour of a machine, would not all of those be frustrated? Our minds and hearts are tuned to the frequency of the divine, however weak our signals may be. Do we really think AI will improve the connection? Social media has already weakened it, through the Scylla of distraction and Charybdis of behavior modification. I suspect AI may weaken it further. Maybe that’s not a bug but a feature.

The imago dei includes both the intellect and love. But they are so intertwined, it is impossible to impact one without the other. Aquinas beautifully describes our intellectual souls as images of the Trinity, “in which is found a procession of the word according to the intellect, and a procession of love according to the will.” Our minds and hearts cannot be separated just as the Trinity cannot be separated. The Christian ideal is to think with our heart (cogitationes cordis - Ps. 33, Sacred Heart Solemnity introit). It’s what Mary did when she continually “pondered these things in her heart.”

Yes, our minds are imperfect. So, God supplies us with many graces to be conformed to the mind of Christ. Saint Paul warns us that the unspiritual will see this as foolish.

May we be fools in the eyes of the world, and unite our mind and heart with our Lord’s, and not with a machine.

My (inadequate) summary of his argument: (i) AI is defined as the capacity of computers to simulate intelligent behaviour (not actually be intelligent). (ii) Only a living thing can actually be intelligent. (iii) But in reality, AI is actually designed to fool us into thinking it is truly intelligent. It does this by producing words which are “exclusively a manifestation of intelligence.” (iv) Therefore, AI is duplicitous. When AI produces words, the “speech event is make believe, an illusion, with nothing below the surface or behind the curtain. It is curtains all the way back.”

Christian tradition holds that only angels (including fallen ones) and humans have been given the gift of consciousness by God. If a machine, like AI, were to actually possess consciousness, it might mean that something is actually possessing it. And “alignment” does not work with that something (ask an exorcist). This is way beyond my pay grade (which was $5.25 per month in prison), so I’m not going to open this can of worms here. Besides, others have written about the supernatural risks of digital idols.

All Christian-themed writing at the Abbey is free to read for all. If you would like to make a donation, or take out a paid subscription, it will be very gratefully received.