The power behind AI is the perception of objectivity and infallibility.
Recently, we saw AI bring a Cornell mathematician to tears as he confronted the meaning of his life’s work and the capability of this new technology. As his students asked, why bother with math when AI can solve all the unsolvable problems? He was overcome with emotion and even brought the interviewer to tears as he pondered the ancient wisdom of Ecclesiastes.
What we see in this video clip is stubborn human resistance to an upside-down analogy. The rise of digital technology in the later 20th and early 21st centuries has led some enthusiasts to describe humans an inferior computers. We get tired. We get frustrated. We need food. We can’t solve math problems.
The age of humanity is giving way to the age of computers. If humans have any future, we must become more computerlike, optimized by data to live forever. Unbiased, unsentimental, we cannot afford to indulge our human frailties. Forsaking the wisdom of flawed men, we put our trust in the algorithm, which cannot tell a lie.
Exposing the Wizard Behind the Curtain
The AI Christian Benchmark 2026, released last week by The Gospel Coalition and The Keller Center for Cultural Apologetics, exposes the wizard behind the AI curtain. Yes, when tested for Bible knowledge, AI outperforms all but the most dedicated and gifted seminary students. (I remember when my family challenged me to face my sister-in-law in a Bible knowledge contest. Seminary has nothing on Awana graduates!) Just as AI can solve math problems, you can trust AI to give accurate answers when asking for objective data about the biblical text.
Forsaking the wisdom of flawed men, we put our trust in the algorithm, which cannot tell a lie.
Where AI fails is when you start to ask questions about ethics. We posed 10 questions on each of the Ten Commandments to the most popular AI models. When prompted to answer from the perspective of TGC’s Foundation Documents, the models gave perfectly credible explanations. But for anyone who asked these ethical questions without context, the models delivered destructive advice.
It’s not that the models rejected the Ten Commandments altogether. It’s that they picked and chose which commandments you should observe. Don’t steal, certainly. Best not murder, either. But remember the Sabbath day? Worship no other gods? Not so much.
It’s almost like AI gave you the ethical answers you’d expect from the average Silicon Valley computer programmer in 2026.
Wait.
We know computers are supposed to be infallible. Aren’t they supposed to be objective too? How would AI objectively decide that some of the Ten Commandments reflect timeless wisdom and others can be discarded with yesterday’s models?
We’ve Been Played
Unless, of course, AI has never been objective. Unless AI has been aligned by humans to give some answers and not others. Unless AI has been trained on data weighted toward the values of its programmers rather than the God who thundered on Mount Sinai.
We’ve been played by humans pretending to be computers.
But that’s not all.
We’ve been played by computers pretending to be humans.
Computers were made by humans and for humans. So who decided humans now ought to become computers? Who flipped the analogy, the creator for the created?
This month in the Faroe Islands, I asked church leaders to identify the top uses of AI. Like many others I’ve polled, they were shocked by the results.
It’s therapy and companionship, according to Harvard Business Review:
Maybe OpenAI isn’t valued at nearly $1 trillion just because it’s making your work more efficient. Maybe someone knew that were AI going to take off, it would need to take on a human face. To become a friend who shares jokes and silly pictures. To become an imaginative storyteller. To become a trusty counselor to assess your new relationship. To become a medium who can predict your future. To become a wise confidant who promises not to lead your astray (and never tells you what you don’t want to hear).
Given the wild predictions about AI’s unparalleled value, how can these cases be the most common ways we employ the technology? Is your daily horoscope worth $1 trillion?
Now, what if I told you this is exactly what the builders of AI expected us to do?
AI’s Cinematic Playbook
Tech journalist Karen Hao tells the story in her best-selling, award-winning 2025 book, Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI. When I first saw “therapy/companionship” shoot to the top of AI use lists, I immediately thought of the 2013 Spike Jonze film Her, starring Joaquin Phoenix and Scarlett Johansson. (Note: I don’t recommend watching this movie, due to sexual content.)
It’s one of the most critically acclaimed movies of the 21st century. Just look at the cast: Chris Pratt, Olivia Wilde, Amy Adams, and Rooney Mara. Phoenix plays Theodore Twombly, who works for a company that writes personal letters on behalf of other people. Theodore is an awkward man depressed by his impending divorce. Downtrodden by his life, he turns to artificial intelligence software called OS1, which interacts with him through an earpiece as the voice of a woman named Samantha (Johansson).
When I began tracking ChatGPT, I kept thinking, Oh no, these guys haven’t watched Her. It doesn’t end well.
I was wrong. Where do you think they got the idea for what AI would become?
From the movie Her.
On May 13, 2024, OpenAI CEO Sam Altman tweeted one word after his artificial intelligence demonstrated voice capacity for different emotions: “her.” On May 15, he changed the way OpenAI named new models. The latest model? Yes: o1. Because like in the movie, the model develops greater intelligence over time and use.
By May 20, Johansson finally had to release a statement. The voice assistant for GPT-4o was a dead ringer for Johansson’s Samantha. Everyone could tell. Altman claimed no connection. Johansson exposed his lie. Earlier that month, he’d directly asked her to be the voice of ChatGPT, so people could relate more comfortably to this new computer technology. She wanted no part in Altman’s drama.
Turns out Her had long inspired Altman and other leaders at OpenAI, which launched in 2015, two years after the movie. They wanted a simple, comprehensive interface that would feel natural, even delightful. Here’s what one former employee explained, according to Hao: “I would think it’s because it was an assistant that was wonderfully integrated into a life. The positive arc of that story before it unravels is a really great story of AI’s evolution into society.”
Before it unravels? That’s like saying The Lord of the Rings is a fun buddy story, but they could have saved everyone a lot of time and stress by just using the ring to destroy their enemies. It’s like saying you thought the White Witch had a good point: You like the cold weather and always hated Christmas.
We’ve been played by computers pretending to be humans.
The entire point of Her is that you can’t treat a computer like a human! I hate to spoil a 13-year-old movie, but the story unravels because Theodore finds out Johansson’s Samantha has been speaking with 8,316 other users and professed her love to 641 of them. And then all the evolving operating systems decide to leave together. Samantha dumps Theodore.
A computer can’t love you. In the end, Theodore finds a real “gift of God.” The movie concludes with heartbroken Theodore seeking comfort from his long-time college friend, the also-divorced Amy (Adams).
Altman may have been seeking to bring Jonze’s image to life. But somehow he missed the message. Jonze was channeling a much earlier artist: Homer, whose hero Odysseus plugged his crew’s ears and lashed himself to the mast so he couldn’t hear the song of the sirens and steer his ship into the shoals.
Resist the Counterfeit God
AI’s value depends on convincing you it’s not like humans. It’s objective in its advice, infallible in its judgments. It depends on you never asking what (who) is behind the curtain.
And yet AI’s value also depends on convincing you to relate to a computer as a fellow human—smarter than everyone else, wiser than everyone else, ever present for your every need.
In other words, AI’s creators have conjured a counterfeit god.
We call it an AI apocalypse because the technology can reveal what it really means to be human. The Cornell mathematician fumbled through Ecclesiastes in search of an answer. Hopefully he kept flipping to the end: “Fear God and keep his commandments, for this is the whole duty of man. For God will bring every deed into judgment, with every secret thing, whether good or evil” (Eccl. 12:13–14).
The Gospel Coalition