Part 1 of 2 · 15 November 2026 · 9:19 pm AEDT

I’ve noticed a quiet competition in some conversations about AI. Someone mentions a skill they picked up with Claude. Someone else tells you what they learnt with ChatGPT. I’m interested in what passes between them in that moment. Each is telling a story about what they can now do, but also naming a source of knowledge they’re willing to be seen trusting.
Imagine two families leaving the same school open day. Both like the idea of an AI tutor that can explain a maths problem to a child who is embarrassed to ask again. One parent also welcomes the school’s plan to use the system to recommend next year’s classes. A teacher might overlook a quiet student, she says. Another parent asks what happens when the system overlooks one. Who can challenge the recommendation, and what evidence would the school accept?
A child can tell you whether an explanation helped. But if a prediction keeps them out of advanced maths, they may never get the chance to show what they could have done. That’s where I think trust starts to become authority. The school has given the prediction a say in what the child gets to try, and therefore in what anyone will ever know about their ability.
I think of faith as a further step. It takes hold when confidence in a tool becomes a conviction about the future it promises. In 2008, Robert Geraci traced parallels between Jewish and Christian visions of life beyond death and the promise that we might live on by transferring our minds into machines. In Artificial Religion, Mark Coeckelbergh looks at how religious stories shape the hopes we invest in technology. Their work makes me wonder what happens when one of those hopes becomes something people organise their lives around.
A promise can bind a community long before anyone can test it. The story behind it could even be made up. Imagine a group that comes to regard digital immortality as a certainty, or to see handing decisions over to AI as a moral duty. If belonging depends on accepting that belief, asking for evidence might be heard as a refusal of loyalty. People who see the same promise as a threat might build communities and institutions of their own.
That possibility raises a question about who gets to speak for the machine. Imagine a community that treats a model’s advice as a guide to living well. An update reverses its answer to a question the group considers fundamental. Some members might accept the new answer as a deeper insight. Others might preserve the old one and accuse the company of corrupting it. Each faction could claim fidelity to the same authority. The split might turn sectarian if each side came to see the other’s interpretation as a betrayal of a belief they once shared.
My hunch is that, in some societies, such commitments could one day tell us more about who stands with whom than religious affiliation does. Some movements around AI might even become religions in their own right, with rituals, authorities and rules about what it means to live well. I keep coming back to that small moment in conversation. How much would have to change before “Which AI do you trust?” starts to sound like “What do you believe in?”
The idea and argument are my own. I used ChatGPT to help draft and refine the essay.
Leave a Reply