Why I Built AI for the Church
I have spent most of my working life building technology, and most of my adult life trying to take faith seriously. For a long time those two things lived in separate rooms. One was the world of products, metrics, and machines that get faster every eighteen months. The other was the world of Sunday mornings, of a pastor I trusted, of the long slow work of becoming a different kind of person. I assumed they would stay in separate rooms. They didn't.
When I look at the most powerful technology of our time, I don't see a neutral tool. I see something with enormous gravity — it pulls our attention, our money, and increasingly our trust toward whoever points it. Most of the people pointing it are pointing it at commerce: at getting us to buy more, scroll longer, want things we didn't know we wanted. I kept asking a stubborn question. What would it look like to point this same power at the local church? Not to make the church more like a tech company, but to give the church back the one thing technology has been quietly taking from all of us — time and attention for what actually matters.
That question is why I built SoapBox.
The shepherd and the spreadsheet
Spend time around pastors and you learn a quiet truth: most of them did not enter ministry to run an organization. They were called to care for people. But somewhere along the way the calling got buried under the operation. Newsletters. Volunteer schedules. Sermon prep with a deadline that arrives every seven days, forever. Tracking who showed up, who slipped away, who is going through something and hasn't said so. The administrative weight of a modern congregation is enormous, and it falls hardest on the very people we most want freed up to be present.
This is where I think technology has an honest and humble job to do. Not to be wise. Not to be holy. To carry the load that was never sacred in the first place.
The machine carries the spreadsheet so the shepherd can carry the sheep.
That sentence has become a kind of north star for how we build. When a tool drafts the routine email, organizes the follow-ups, surfaces the person who has gone quiet for three weeks, it is not replacing pastoral care. It is clearing the desk so pastoral care can happen. The goal is a pastor who ends the week with more margin for the hospital visit, the hard conversation, the prayer that can't be rushed — and less time staring at a screen doing work a machine should have done.
[Alan — a 2–3 sentence personal moment fits well here: a specific pastor or conversation that made the "buried under the operation" problem real to you.]
The person awake at 2 a.m.
There is a second reason I built this, and it has nothing to do with productivity.
The spiritually curious do not keep office hours. The questions that matter most rarely arrive on a Sunday at 10 a.m. when the building is open and someone is available to talk. They arrive at 2 a.m., in the dark, when a person is afraid or grieving or simply wondering whether any of this is true and has no one to ask. For most of human history, that person was alone with the question until morning, if they made it to morning still wanting to ask.
I do not believe a machine can be a friend, a pastor, or the Spirit of God. I want to be clear about that, because it is the exact place where this technology becomes dangerous if you forget it. But a machine can do something real for the person awake at 2 a.m. It can listen without flinching. It can open the Scriptures with them. It can answer an honest question honestly, without shame, and — this is the part that matters most — it can point them toward a real church, real people, a real community where the actual work of faith happens. The aim is never to keep someone talking to software. The aim is to walk them to the door of a church and let them go in.
If even one person who would otherwise have stayed silent in the dark instead takes a first real step toward faith, the entire effort is worth it.
Carrying Scripture across languages
The third reason is the one I feel most personally. I was not born into the English-speaking, Sunday-morning world that much of American Christianity takes for granted. I know what it is to stand a little outside the language, to need a bridge.
The good news of the gospel was never meant to belong to one language. And yet, in practice, so much of the digital church has been built for English first and everyone else later — or never. We have built SoapBox to carry Scripture and spiritual conversation across more than 140 languages, so that the person seeking God in Farsi or Tagalog or Swahili is not a second-class member of the family but is met in their own tongue. There is something deeply right about technology being used to widen the door rather than narrow it.
[Alan — a 2–3 sentence personal moment fits well here about language, your own background, or carrying faith across a cultural line.]
The temptation, and the line we will not cross
I would be lying if I said this work has no temptation in it. The same tools that can free a pastor can also flatter people into thinking the machine is wise. The same AI that can answer a midnight question can be tuned to keep someone scrolling instead of sending them to a real community. The line between serving the sacred and quietly replacing it is thinner than the industry wants to admit.
So we have drawn the line on purpose, and we keep redrawing it. Technology should serve the sacred, never stand in for it. Anything we build is meant to return a person to people — to a pastor, a congregation, the slow embodied life of faith that no software can simulate. As the Scripture puts it, "faith comes from hearing, and hearing through the word of Christ" (Romans 10:17, ESV). Our job is to help that hearing reach further, not to pretend to be the voice.
That is the whole of it. I built AI for the church not because the church needs to be more like the machine, but because the machine, used with restraint and conviction, can give the church back its time, its reach, and its attention — so that the people doing the sacred work have more of themselves left to do it.