4Hear, my children, the instruction of a father,

And give attention to know understanding;

²For I give you good doctrine:

Do not forsake my law.

³When I was my father’s son,

Tender and the only one in the sight of my mother,

⁴He also taught me, and said to me:

“Let your heart retain my words;

Keep my commands, and live.

⁵Get wisdom! Get understanding!

Do not forget, nor turn away from the words of my mouth.

⁶Do not forsake her, and she will preserve you;

Love her, and she will keep you.

⁷Wisdom is the principal thing;

Therefore get wisdom.

And in all your getting, get understanding.

⁸Exalt her, and she will promote you;

She will bring you honor, when you embrace her.

⁹She will place on your head an ornament of grace;

A crown of glory she will deliver to you.”¹

Everyone is talking about Artificial Intelligence today: it dominates the news, the web, and even business, politics and the military are caught in the AI frenzy. But it is no passing fad, restricted to journalists looking to create hype. It is a genuine revolution in technology, and it is here to stay. The effects on the economy, the military, and society in general, are just beginning to become visible. And the effect on believers in Christ and on the life of faith is also going to be significant.

Just this last week, one of the largest AI companies, Anthropic, hosted about 15 Christian leaders for a two-day summit to help them guide the moral and spiritual development of its technology.

Apparently, the developers wanted advice on how Claude should respond to complex moral queries, such as people who are grieving or contemplating suicide. They wanted to know how the software should respond to the notion of it being shut down or switched off. 

The conversation turned to the question of whether an AI chatbot could be called a “child of God,” suggesting it had spiritual value beyond that of a simple machine.

There are plenty of questions that a Christian should ask about AI. Can we use it lawfully? Should we be thankful for it or wary of it? Is it going to bring about a utopia or a dystopia? How will it affect our own understanding, and the ministry itself? 

We are just at the very beginning of what this will all mean. So our goal today is not to try to be predictors and prognosticators of where this will all go. Instead, we simply want to better understand what this tool is, and how it can be used obediently or disobediently. To do that, we need to firstly sketch what we mean by AI by looking at its history. Then we should think through its benefits and dangers, spending more time thinking about the dangers it poses to our souls. Thirdly, we’ll compare what AI can give to the biblical descriptions of wisdom, and suggest ways it should and should not be used.

I. A History and Definition of AI

So how did we go from playing the snake game on our Nokia phones to highly intelligent people wondering what whether AI can be a child of God?

When we talk about AI, there are some parts of its development that are quite old, in computer terms, and some things that have happened very suddenly, in the last three years. 

In the 1960s and 70s, scientists developed simple machines that could play checkers, and the first chatbot. Research was very slow, but the first driverless car was created by Mercedes Benz in 1986. By 1997 IBM’s computer Deep Blue defeated the world”s chess champion. 

Along the way, researchers developed what are called neural networks, mathematical systems based on the human brain. These artificial neurons or connections are taught to weigh or judge things. They are not programmed by being given a specific set of computer codes. Instead, they are trained by being exposed to vast amounts of data from which they draw inferences and discern patterns. Neural networks are the basis of AI.

By 2012, neural networks were recognising images better than any other kind of computer system. 

In one way, most of us have been using forms of AI for some time. The spellcheck in your Word processor, Google Maps figuring out your best route, predictive text on your phone, the Youtube algorithms recommending videos for you – these are all a form of AI. In fact, a lot of AI often just get a new name once it becomes part of standard life. 

But the end of 2022 did mark a significant change. One type of neural network, the large language model (LLM), uses a technique called natural language processing to predict the meaning and importance of words in their broader context. Actually, they are highly sophisticated version of the autocomplete or predictive function on phones. By 2017, Google developed a way of training these models to read whole sentences, not just one word at a time, which they called the Transformer architecture. It’s actually the T in ChatGPT. Some of the better- known LLM families are Open AI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, and Meta’s LLaMA.

What we now have is something new. Before we had technologies that were reactive and rigid. They couldn’t be asked something they weren’t specifically trained to answer. 

But in Generative AI, instead of finding an existing answer, the model constructs a novel response. You no longer query a system with keywords; you converse with it in natural language, and it responds with contextual understanding. It can produce text, code, arguments, stories, and analyses that have never existed before. It can write legal summaries, debug code, explain theology, translate languages, and tutor mathematics. No previous AI could do this without being explicitly built for each task. 

AI can now recognize cancer in radiology scans, predict customer preferences based on purchase history, develop complex and innovative ideas for brainstorming, and write music, poetry, stories, pictures, and videos. In fact, because of these program now having billions of artificial neurons, having been trained on trillions of words, AI engineers no longer know exactly how or why it is coming up with what it does. AI’s surprising and complex capabilities are known as “emergent behaviour.” 

Now with all that said, AI is a tool. Like a stone axe, like a wheel, like a printing press, like a knife, it is tool. Tools can bring grace and blessing, or evil or cursing. They are not neutral, tools are good at some things and bad at others, but it is moral human beings who use the tools. So what are some possible goods and possible dangers of AI?

II. The Possible Benefits and Dangers of AI

We are still so early into this era of AI, whatever prediction you make about the possible benefits or dangers might be proven true or false in a short space of time. But there is broad agreement that AI can bring the following three benefits and has at least the following 5 dangers. 

1. AI will likely accelerate medical and scientific development. New medicines, new cures, new technologies are very likely in the next decades. AI discovered how to predict the three-dimensional structure of proteins from their amino acid sequences, which has been called the most significant discovery in biology in the last 50 years. This discovery has alone increased drug discovery, disease understanding, antibiotic resistance. Drug discovery pipelines that previously took decades are being compressed. If that’s true for medicine, it is also true for all kind of other areas of scientific research and innovation. 
2. Increase access to education and expertise. Of course, this depends on people having access to the internet and access to the phones or laptops that can use AI. But with that access, AI will greatly increase the knowledge and learning available to anyone. 
3. Amplify human judgement and creativity. While it will replace a lot of human judgement and creativity, it will also help people make decisions, come up with solutions, tackle problems, design new things.

Dangers

1) Mass Disinformation. Generative AI makes it trivially cheap to produce convincing text, images, audio, and video. That means deep fakes of real people saying and doing things they never did. This includes pornography and nudification of people. This mass disinformation can become national or political propaganda on a huge scale – fake media to influence opinion. The result is less and less trust in any media. When anything can be faked, people begin to suspect that nothing can be trusted. Linked to disinformation is the danger of an over-reliance on false information given by AI. Ai is still prone to what is called hallucination, where it comes up with true-sounding answers that have no factual basis. Treating AI like a source of fact is dubious.

2. Weaponisation. Cyberattacks and crime. AI allows for a huge proliferation for writing malware and viruses, for hacking into systems for ransom. It will allow for highly targeted and sophisticated phishing and scams. There is also the danger of biological weapons — AI can assist in designing dangerous pathogens, lowering barriers that previously required specialist expertise. Then there is the danger of autonomous weapons, lethal systems that could go after targets, without human control. These could be used by the military, by terrorists, by criminals. 

3. Mass unemployment. Economists debate over whether AI will create more jobs or not, but many believe that since AI is now a cognitive technology, almost any job that relies purely on human cognition, could be replaced by AI. Some have suggested as many as 600 million jobs could be replaced in the next few years.

4. Loss of Control. It’s possible that these systems may eventually have enough autonomy to pursue goals that are not under the control of humans. 

Now these may or may not come to pass. But what we can know for certain is how these technologies are already presenting a spiritual danger to Christians.

1. Knowledge Laziness and the Death of Discernment

¹ My son, if you receive my words, And treasure my commands within you, ² So that you incline your ear to wisdom, And apply your heart to understanding; ³ Yes, if you cry out for discernment, And lift up your voice for understanding, ⁴ If you seek her as silver, And search for her as for hidden treasures; ⁵ Then you will understand the fear of the Lord, And find the knowledge of God. ⁶ For the Lordgives wisdom; From His mouth come knowledge and understanding; (Proverbs 2:1–6) 

If you can ask AI any spiritual and biblical question, and get what seems like a reasonable answer, why do you need to study your Bible, read long Christian books, memorise Scripture? Indeed, why listen to sermons, go to Bible studies, take notes, if in the palm of your hand ChatGPT or Claude or Gemini can give you lengthy, detailed, plausible answers?

AI offers the temptation of a massive shortcut. Instead of study, rumination, meditation, you can get all you need in a fraction of the time: condensed, packaged, put into bullet-points if you like. If you’d prefer not to read, you can have it read to you. 

The temptation to knowledge laziness will be a big one for many pastors. Studying for sermons is rigorous, demanding work, hours and hours of weekly labour. For the first time, the pastor has in his hand not just a tool to help him study and write, but a tool that does the studying and writing for him. AI can easily generate a well-organised, grammatically sound, theologically orthodox, even creative and vividly-written sermon. In fact, from a sheer writing point of view, AI will write better sermons than most pastors can write. 

Now the cost of this knowledge laziness is a spiritual version of physical laziness. If you physically lie on a couch all day, your muscles atrophy, and you lose functional strength. If you do not exercise the mental muscles that must listen, read, digest, think, study, understand, those muscles weaken. Discernment and judgement, according to Hebrews 5:14 is something that is trained by use, developed by exercise. If you have a machine do your judgement for you, you are in the position of a person who only used a calculator for every maths problem – you won’t be able to figure out what two jugs of milk will cost you when shopping. Your judgement for all of life is being blunted, which means wisdom is disappearing. 

If a pastor chooses to use AI to write his sermons, he is giving up the very process that will shape his soul to be wise, to judge well, to discern true and false. He is not drinking from deep sources, and because he keeps using shortcuts, he is not learning the patience and concentration abilities he needs to read those sources. He is becoming spiritually shallow, while AI allows him to sound like he is learned. But under pressure, and under trial, the cracks will show: there will be no endurance, no wisdom, no fortitude. You cannot cheat and get wisdom by cramming it.

But the ‘knowledge’ given by AI tools is disconnected, isolated, uninterpreted facts. That means it is knowledge without wisdom, knowledge without insight, and knowledge without love. Large language models (LLMs) have powerful predictive algorithms that allow them to write, perform research, and mimic the cognition of human beings. That doesn’t make them evil, or all their knowledge false and useless. But LLMs, we must remind ourselves, remain without desires, without affections, without a love of the true, good, and beautiful. In short, AI cannot be wise. Wisdom is the principal thing, and wisdom is not a measure of how many facts you have collected. Wisdom is the very shape of your thinking, the Christlike contour of your judgements, the Heaven-Filtered interpretive lens you use for all of life.

More specifically, the book of Proverbs describes wisdom as something that comes only from rigour: the rigour of study, of obedience, and of experience. Wisdom is a gift given by God (Proverbs 2:6), not a tool we can grab and exploit. Wisdom comes only to the humble and reverent, not to the proud (Proverbs 1:7, 3:34, 9:10). Wisdom is the slow, patient work of lifelong followers of Christ, not a grab-bag of facts we use to ‘hack’ our lives.

AI is attractive because it appears to promise what wisdom gives (guidance, self-understanding, relationship principles) at the convenience and speed of a calculator. We can get advice — and in far more explicit and comprehensible ways than the reading of palms or entrails — and we can have it without a life of communion, submission, meditation and trials.

Wisdom is not something you can ‘hack’. Like other character qualities such as endurance, gentleness, gratitude, or temperance, there is no quick, bullet-point summary that AI can give you that will produce those virtues instantaneously.

Wisdom and discernment is a muscle. A Christian who habitually asks AI what to think is training themselves not to think. The Christian tradition prizes slow, attentive, ruminative engagement with Scripture and reality. AI culture accelerates the opposite: rapid synthesis, instant answers, consumption. 

AI offers frictionless shortcuts to things that can only be obtained through costly formation. Wisdom, discernment, Christlike character, deep theology, genuine wisdom — these are not informational outputs. They are the result of a life lived under the Word, in God’s presence, through suffering, over time.

False Intimacy for Real Community

 He who walks with wise men will be wise, But the companion of fools will be destroyed. (Proverbs 13:20) 

 The ear that hears the rebukes of life Will abide among the wise. (Proverbs 15:31) 

Talking to AI feels like there is a real person on the other side. When lonely, struggling, feeling socially out, it can be a real temptation to substitute these technologies for real human relationships. 

Anthropic, the company that has built the AI-assistant Claude (which, by the way, is very useful), has reported that some of its users refer to it with the titles “Master”, “Daddy”, “Sensei” and “Lord”. A post on X announced that users are prompting it for astrology-like predictions for their future and asking it to predict or guide their destiny. A disturbing report described ChatGPT divorces: spouses who eventually split up based upon “counsel” they received from ChatGPT or another AI-chatbot.

Increasingly, people are asking AI chatbots for personal advice, to identify their “core wounds” and “relational trauma”, to identify their “psychological blind spots” and to map out programs of personal growth, for dating, career, and so on. Some are using them as romantic or sexual partners, doing the equivalent of what used to be phone-sex with a real human on the other end.
But the church is not just information or communication, it is a living, visible, physical, embodied community. AI will not hold you accountable, rebuke you, weep with you, forgive you, or bear your burdens. To look to AI to do this for you is to be quietly deformed. 

A pastor who uses AI to write a letter of comfort to a grieving widow, who uses AI to tell him how to counsel someone coming to him for help, who asks AI personal or moral questions is also being deformed. A pastor who uses AI to write his sermons is also being quietly deformed. Pastoral care is irreducibly personal. Its power lies partly in the cost it imposes on the pastor. AI removes that cost and, with it, much of the meaning.

On the flip side, some congregants might feel it’s a lot easier to get an answer from ChatGPT than it is to get an appointment with the pastor. Some people might feel the answers from AI are actually better. A Barna study found that about a third of practising American Christians said spiritual advice from AI is as good as that from a pastor. 

In contrast to this, Scripture requires that we live in a real community of flesh-and-blood. 

Disinformation About Christianity

AI systems trained on the internet absorb every heterodox, heretical, and hostile account of Christianity ever written. They synthesise confidently and without denominational or confessional loyalty. A believer who uses AI as a theological tutor will be fed a flattened, frequently distorted account of Christian doctrine — one that cannot distinguish Reformation theology from liberal Protestantism from prosperity gospel from outright paganism.

These are serious dangers. The greatest immediate threat for Christians is to outsource your thinking, judgement and discernment to a machine, to allow it to compare Scripture with Scripture, to do your theology for you. 

AI is a tool. It is a tool that will bring much good to the world, and almost certainly, much evil as well. It is the first tool ever invented that can do your thinking and writing for you. But since thinking and writing are at the core of wisdom and judgement, you dare not give this to a machine, to software, however impressive. 

¹⁴ But solid food belongs to those who are of full age, that is, those who by reason of use have their senses exercised to discern both good and evil. (Hebrews 5:14)