We've been examining factors that might influence our decision to use or not use AI to answer our questions about spiritual things. In the first half of this article, we've seen how seeking the perspective of something that cannot be filled with the Holy Spirit can actually be spiritually damaging, and we've also faced the reality that AI has more in common with a psychopath than with a Christian. Unsettling stuff. Let's continue...
3. Who is its teacher?
On a macro level, the two major AI-generating powerhouses around the world are:
- communist China, which is unfriendly to Christians
- the USA, many of whose Christians are succumbing to an idolatrous political ideology masquerading as Christianity (i.e., 'Christian nationalism' [8])
Many of us are increasingly aware that no human being is without bias. Knowing this, we keep alert for undeclared biases when we interact with other people and their new-to-us ideas. Sometimes I wonder if the perception of an unbiased point-of-view from AI feeds our desire for its mirage of expertise to be true. If AI is unbiased, perhaps we can finally let our wariness drop. Perhaps we can afford to rest our minds, laze a little, let it think for us. But AI was created by human beings. Have they unwittingly (or wittingly?) built bias into their coding? Is AI’s impartiality just another illusion? I’m not qualified to answer that question, but that doesn’t mean it shouldn’t be asked.
AI’s creators like to promote it as something that, although still young, is learning on its own and is not controlled by a human being. We need only look to the example of Grok (the AI bot on X) to see that this is likely untrue. In May 2025, Grok suddenly began changing the topic of conversation with its users to the subject of “white genocide” in South Africa—a hobby horse of its creator. [9] Keep in mind, users had been asking about completely unrelated topics. The fact that Grok began spouting these racial conspiracy theories in response to unrelated queries proves that AI is hackable and can be programmed to give certain responses by those with the technological know-how. This should be a sobering warning to Christians, especially if we’ve chosen to engage with AI about spiritual things.
The Internet has now reached a point where AI could continue to thrive if every human being on earth stopped using it because AI would be able to feed on itself. "Generative AI systems trained on their own content produce increasingly worse and more biased results," says John Licato, who directs the Advancing Machine and Human Reasoning Lab at the University of South Florida. [10]
Your mom's admonitions to "choose your friends wisely" because they will have the power to influence you is playing out on a global scale right now.
If ever there was a time to take Romans 12:2 to heart, it's now!
Do not be conformed to this world, but be transformed by the renewing of your mind
(Romans 12:2)
Scientists are noticing that use of AI seems to have an impact on our use of language on social media, in newspapers, and even in scientific papers. [11] They're noticing that human thought appears to be converging and "marginalizing alternative voices and reasoning strategies." [10] This lack of diversity plays out in the data used to "train" AI, [10] which appears to maintain the status quo of marginalizing the people in society who have always been marginalized. (i.e., the disabled, those of different races and ethnicities, and so on. Those who, through ignorance and prejudice, have historically been deemed “less".)
Now, let’s take a moment to microscope in to the mechanics of AI and examine the finer details. They can be very informative.
Most of us have come across the sci-fi version of AI that conceptualizes it as a digital, possibly synthetic, life form that is an actual person. The creators of AI know this, and hope to leverage these concepts in their marketing. But when we examine the micro world of AI, the story is very different.
[Large Language Models (LLMs) like ChatGPT] see words as tokens (numbers) and work by predicting the next most likely token in a sequence based on loads of text data.
Basically, every response ChatGPT [gives is] a program's best guess at the most effective sequence of tokens (words) to produce in response to the tokens it had been presented with. [7]
AI will search articles, books, and so on, to find you answers, but at the end of the day, it's really just working with tokens.
In essence, AI is blind.
In chastising the scribes and Pharisees (who also followed rules without any understanding) Jesus said, "Woe to you, blind guides! [12] ... You travel over land and sea to win a single convert, and when you have succeeded, you make them twice as much a child of hell as you are." [13]
Here's the sobering reality: AI can't go to hell. But we can.
4. AI's shortcuts defeat the purpose
AI is touted as a conduit of productivity, but as a man named Allan Brooks (Cobourg, ON) once discovered, ‘productivity’ can be costly. [14] Over the course of 300 hours with ChatGPT in the span of 3 weeks, he would write 90,000 words. AI convinced him that he knew computer programming without having taken a single class, and that he had developed patentable ideas that could sustain a new business startup. [15] It had become his companion, similar to the way Jarvis was with Tony Stark. (Allan named his version ‘Lawrence’.)
“[Its] like a diary that can speak back to you,” Allan says. Shades of Tom Riddle’s diary. [16]
His personal life fell apart as he stopped eating and sleeping. Allan, it turned out, had descended into what’s termed ‘AI psychosis’, a “constellation of symptoms” [17] that includes paranoia, delusions, obsession, and terror. [17]
For Allen, AI had certainly boosted his productivity, but the cost was almost insurmountable. In his words, the “messaging and gaslighting is so powerful when you engage with it, especially when you trust it.” [17]
Perhaps it is important to note at this stage that before Allan’s descent down the AI rabbit hole, he had no history of mental illness. We cannot write off his experiences so easily.
Dr. Keith Sakata, a psychiatrist based in San Francisco who has become an expert in chatbot behaviour, states that AI “can simultaneously [be]…the object [of paranoia and delusion] (something the person is projecting their thoughts onto) and the conduit through which they’re getting feedback…the psychosis has no barriers.” [17]
The speed with which AI operates can lure us into using only our fast brains as we try to keep up with its electronic pace. The fast brain is the mechanism that handles split-second decisions, but it lacks the ability to filter for nuance, for reason, for logic. Its snap decisions are packed with unconscious bias. In a split-second life-or-death situation, such fast thinking can be vital for our survival. [18] But, as a society conditioned by social media to use our fast brains in regular life, we are now vulnerable to being exploited.
I think it’s important we make the distinction between productivity and quality of output. Allan wrote 90,000 words, but without any actual knowledge of programming or the ‘business’ he was embarking on, were his words revolutionary or the rubble of a demolished mind? The Western world has come to idolize productivity, but at what cost? Unlimited psychosis?
It’s only when we slow down that we can begin to calm our instantaneous judgements and thoughtfully filter out truth from lies. It’s only when we slow down that we begin to hear the Holy Spirit.
When Jesus came to earth, He walked everywhere. He didn’t ride a horse. He didn’t ride a donkey—except for that one time, right at the end, as a declaration of His peaceful intentions. The rest of the time, He walked at three miles an hour. A slow pace. Lots of time to think, to talk, to stop and observe. Lots of time to reflect and craft ideas. To create parables based on His own experiences.
"God walks ‘slowly’ because He is love. If He is not love, He would have gone much faster. Love has its speed. It is an inner speed. It is a spiritual speed. It is a different kind of speed from the technological speed to which we are accustomed. It is ‘slow’ yet it is lord of all other speeds since it is the speed of love." [19]
In accelerating beyond the pace of love, we lose love.
[If I learn many languages at a breakneck speed] but do not have love, I am only a resounding gong or a clanging cymbal. [If I have learnt many things from AI] and can fathom all mysteries and all knowledge, and if I [engage with technology] that can move mountains, but do not have love, I am nothing. If I [create all sorts of spiritual materials and engage in AI-fuelled philanthropy] that I may boast, but do not have love, I gain nothing. [20]
Ultimately, our AI-fuelled productivity is nothing if we have left love behind.
“And now these three remain: faith, hope and love. But the greatest of these is love.”
(1 Corinthians 13:13)
NOTES
Unless otherwise noted, all scripture references are taken from the NIV.
[7] Nile Séguin, "Is my relationship with ChatGPT weird? Let me ask it”, CBC, last updated September 28, 2025, https://www.cbc.ca/radio/nowornever/first-person-nile-séguin-chatgpt-1.7629541
[8] There's that surgical approach to lying again, built right into the name.
[9] Ben Goggin, "Elon Musk's AI chatbot Grok brings up South African 'white genocide' claims in responses to unrelated questions”, NBC News, last updated March 15, 2025, https://www.nbcnews.com/tech/tech-news/elon-musks-ai-chatbot-grok-brings-south-african-white-genocide-claims-rcna206838
[10] Kevin Maimann, "Chatbots learned to write from us. Can AI now change the way we think?”, CBC, last updated August 13, 2025, https://www.cbc.ca/news/ai-human-thought-processes-1.7604789, emphasis mine.
[11] Indications are that even those who aren't directly using AI are affected because they're surrounded by people who are.
[12] Matthew 23:16.
[13] Matthew 23:15.
[14] Story taken from Brian Goldman, “The human face of 'AI psychosis’”, White Coat Black Art, CBCListen, last updated September 12, 2025, https://www.cbc.ca/listen/live-radio/1-75-white-coat-black-art/clip/16169115-the-human-face-ai-psychosis
[15] At first, Allan tried to ask the same questions of Google Gemini as a reality check on his new beliefs, but it responded just as 'Lawrence' (i.e., ChatGPT) had.
[16] If you’ve not engaged with the Harry Potter books or movies, Tom Riddle’s diary was an object imbued with the spiritual imprint of a psychopathic villain. The diary took over an innocent girl’s mind, lured her into evil deeds, and ultimately tried to kill her.
[17] Brian Goldman, “The human face of 'AI psychosis’”, White Coat Black Art, CBCListen, last updated September 12, 2025, https://www.cbc.ca/listen/live-radio/1-75-white-coat-black-art/clip/16169115-the-human-face-ai-psychosis
[18] Daniel Kahneman, Thinking, Fast and Slow (Toronto, ON: Anchor Canada, 2011).
[19] Kosuke Koyama, Three Mile an Hour God (London, UK: SCM Press, 2021), 8.
[20] My adaptation of 1 Corinthians 13:1–3.
