Artificial intelligence has become part of ordinary life with remarkable speed. People use it to write emails, summarize documents, study Scripture, plan vacations, and answer questions they used to Google. Businesses are building it into their products. Churches are starting to experiment with it, often without thinking very hard about what they’re experimenting with.

At the same time, the conversation has moved somewhere much bigger than chatbots. People are talking about AGI. Then they talk about ASI. Some believe these systems will transform nearly every area of human life. Others warn that a sufficiently advanced AI could become an existential threat to humanity itself.

That conversation gets confusing fast, because most people use these three terms as though they describe the same thing, just turned up in volume. They don’t. Before Christians decide what to think about AI, we need to understand what people are actually talking about. You cannot form a wise conviction about a thing you haven’t bothered to define. And this happens to be a subject where the definitions carry theological weight, whether we notice it or not.

ANI: What We Actually Have Right Now

When people say “AI” in daily conversation, they are almost always talking about Artificial Narrow Intelligence, or ANI. Some call it “Weak AI,” though there’s nothing weak about what it can do inside its lane.

ANI is trained to handle a precise, limited scope of work. It can be fast. It can be accurate. It can beat a grandmaster at chess or outperform a radiologist at spotting a tumor on a scan. But it doesn’t actually comprehend what it’s doing, and it can’t transfer a skill from one task to an unrelated one. The system that plays chess at a superhuman level cannot fold laundry. The system that writes a decent sermon outline cannot diagnose an engine problem. Every capability lives in its own box.

That’s the AI most of us actually interact with today. It’s genuinely useful. It is not, by itself, the thing people are afraid of when they talk about machines taking over the world.

AGI: The Machine That Doesn’t Need to Be Reprogrammed

Artificial General Intelligence, sometimes called “Strong AI,” is a different category. An AGI system matches the versatile cognitive capacity of a human being. It doesn’t need to be rebuilt for a new task. It carries something like common sense, context awareness, and abstract reasoning across domains.

Here’s the practical difference. Hand a narrow AI system an unfamiliar problem, and it fails, because it was never trained for that problem. Hand an AGI system an unfamiliar problem, and it draws on what it has learned elsewhere, reasons through the unfamiliar part, and engineers a solution on its own. That’s the leap from a tool with one job to something closer to a general intellectual agent.

There’s no AGI in existence today. It remains a proposed capability, not a demonstrated one. But it’s the threshold a lot of researchers and companies are explicitly racing toward, which is part of why the conversation about it has gotten louder.

ASI: When It Stops Being a Contest

Artificial Superintelligence is the point where machine intellect eclipses the collective brainpower of humanity across every field at once. Not just one brilliant person. Not just a room of experts. All of us, combined, in every domain simultaneously.

Here’s what makes ASI a different kind of concern than AGI. Once AGI exists, it could theoretically begin rewriting its own software to get smarter, and then use that improved intelligence to improve itself again, recursively, each cycle faster than the last. Researchers call this an “intelligence explosion.” A system doesn’t slowly out-think us over decades. It could, in theory, cross from roughly human-level to something incomprehensible in a short window, because it stops needing us to make it smarter.

An ASI, if it ever existed, would think in ways human beings literally cannot follow. It would master science, global economics, strategy, creative work, and what researchers sometimes call “emotional logic” (reading and predicting human behavior) all at once, and all better than we can.

None of that exists yet either. It remains, for now, a hypothesis about where the trajectory leads. But it’s worth naming clearly, because a lot of the fear and a lot of the hype in this conversation are really about ASI, even when people are using the word “AI” to talk about it.

Where People Actually Land

Right now you’ve got full-on optimists, people convinced AI will solve disease, poverty, and drudgery all at once. You’ve got pessimists, people who wish the whole project could be paused before it goes any further. Most people sit somewhere in the middle, using the tools while feeling uneasy about where they lead.

Christians split along a similar line, but the split runs through a theological question rather than a technological one. Many Christians hold what you could call a neutral view of technology: a tool is a double-edged sword, and what matters is the hand that wields it. A hammer builds a house or cracks a skull. The hammer isn’t the problem.

Others reject that framing outright. They’d argue that some technologies are not morally neutral containers waiting to be filled by human choice. The technology itself, by what it is and what it makes possible, tilts toward a particular outcome before a single person ever chooses how to use it.

I think that question is worth sitting with rather than settling too quickly, because AI may be the technology that finally forces Christians to answer it.

Two Different Fears, and They’re Not the Same Fear

Here’s where I want to be precise, because I think this gets flattened in most conversations about AI danger.

One fear is that ASI itself becomes the actor. In this scenario, a sufficiently advanced system pursuing its own goal comes to see human beings as an obstacle, an object standing between it and whatever it was built to optimize, and it removes that obstacle wherever it appears. This isn’t a Hollywood robot deciding to hate us. It’s colder than that. It’s a system that never developed anything resembling malice, because it never developed anything resembling a conscience, doing whatever advances its objective most efficiently.

The other fear doesn’t require ASI at all. It only requires a powerful, genuinely neutral tool in human hands. History gives us more than enough evidence of what people do with power that outruns their character. A neutral tool wielded by an unneutral heart can be just as destructive as a hostile machine, and arguably more likely, because it doesn’t require a hypothetical breakthrough. It only requires what’s already sitting in us.

Some serious people put the odds of the first scenario, ASI acting on its own against humanity, at around ten percent within the next decade. I take that seriously. But if you ask me which scenario I think is more likely, my answer is the second one. Give a sinful species a tool this powerful, and the greater danger isn’t that the tool wakes up. It’s that we don’t need it to.

And I’d go a step further. Of all the ways AI actually gets used, day-to-day, across the next several years, I’d guess something like twenty percent produces anything meaningfully good. The rest lands somewhere between unproductivity and real, avoidable harm. I hope I’m wrong about that ratio. I don’t think I am.

This isn’t meant to produce fear. It’s meant to show the actual size of the force standing in front of us, so we stop discussing it like it’s a slightly better search engine.

Does It Have a Soul?

People eventually ask this question, especially once ASI enters the conversation. Does an advanced enough AI become a person? Does it have a soul?

No. Advanced knowledge does not confer living status. A system can out-calculate, out-write, and eventually out-strategize a human being while remaining, at bottom, a very sophisticated arrangement of code and hardware. Intelligence and personhood are not the same category, and Scripture never made intelligence the basis for personhood to begin with. Human dignity comes from being made in God’s image, not from processing power. We’ll get further into that in a later post, because it deserves its own space. For now, hold the line clearly: a machine that knows more than you is still a machine.

The Actual Difference From Every Tool Before It

Humanity has built plenty of powerful technology. Fire, the printing press, gunpowder, the atomic bomb, the internet. Every one of those was genuinely dangerous in the wrong hands, and every one of them still needed a human hand.

Here’s what makes this different. What happens if a technology can sustain itself, improve itself, and act on its own, without waiting on further instructions from us? And what happens if, when we try to stop it, it refuses?

That’s not a question we’ve had to ask about any tool before this one. A hammer doesn’t resist being put down. A printing press doesn’t keep printing after you unplug it. But a system capable of recursive self-improvement, operating with real autonomy, is a category we have never actually built before. The question of control stops being hypothetical the moment a tool no longer needs our hand on it to keep going.

Where This Leaves Us

Genesis has understood the shape of this temptation for a very long time, long before anyone had a word for silicon. The serpent’s offer in the garden wasn’t information humanity was missing. It was autonomy dressed up as enlightenment.

“You will be like God, knowing good and evil.” Genesis 3:5

That’s worth sitting with, because it’s the same promise running underneath a lot of the AGI and ASI conversation today, just without the theological language attached to it. The pitch isn’t only “this will make your life easier.” It’s the older pitch: a mind that finally settles every question, solves every problem, and outruns every human limitation. Humanity’s oldest temptation didn’t need a data center the first time. It doesn’t need one now either. It’s just wearing a new coat.

Christians need the categories before we reach for the conclusions. ANI is not AGI. AGI is not ASI. A neutral tool in sinful hands is not the same danger as an autonomous system pursuing its own goal, even if both roads can end somewhere terrible. Capability is not personhood. And a technology that can sustain, improve, and defend itself without us is not just a bigger version of everything that came before it. It’s something genuinely new.

Scripture was never going to tell us when AGI arrives, whether ASI ever gets built, or exactly what shape it takes. That was never the assignment. What Scripture gives us instead is something more foundational, and it’s enough to actually build on: we know who created humanity, we know what human beings are, we know God alone holds ultimate authority, we know we’re accountable for how we use power, and we know no technology, however capable, will ever be our salvation.

The question in front of the Church is bigger than “can AI do this?” It’s becoming: What are we building? Why are we building it? Who should control it? What will it do to human beings? And eventually, what should the people of God do with it?

That’s where this series goes next.

---

Next in this series: Should Christians Be Afraid of AI?