The Sequence Problem:
Why When You Use AI
Matters More Than Whether You Use It.

Most of the AI debate in churches is about the wrong question. The question is not whether pastors should use AI. 64% already do. The question is when in the workflow they open it.

That distinction — the sequence — is the difference between a tool that sharpens your theology and a tool that replaces it. And almost nobody is talking about it.

The Monday Rule

The Gospel Coalition published a guardrail so simple it fits in one sentence: “Begin Monday with prayerful reading of your text, allowing the Spirit to speak before engaging any tools.”

That is the line. Not which tools you use. Not how many. When you open them.

If you sit down with the text on Monday morning — no commentary, no AI, no helps — and wrestle with it until it starts to wrestle back, then AI becomes a research assistant. It confirms cross-references. It surfaces historical context you might have missed. It organizes the evidence you've already encountered in your own study.

But if you open ChatGPT before you open your Bible — if the first encounter with Sunday's passage is through a machine-generated outline — something fundamentally different has happened. The AI did not assist your study. It replaced it.

The Master's Seminary Got Specific

The Master's Seminary published a theological framework that categorizes sermon preparation into four sacred dimensions: worship, sanctification, service, and artistry.

Sermon prep is “an act of worship” and “an encounter with the living God.” When AI generates the outline first, the pastor loses the worshipful encounter with the text. It is also a means of sanctification — “drinking in the Word results in spiritual growth” — and AI bypasses that transformative process entirely.

Their conclusion is blunt: treating sermon prep as content production rather than sacred duty “declares the ends justify the means.”

9Marks sharpened this further. A pastor who “turns to various helps before wrestling with a biblical text short-circuits his understanding of the text and the sermon that results.” The failure to labor personally with a text “can easily turn preaching into a rhetorical performance.”

“ChatGPT does not feel the heartache of the couple struggling with infertility. Gemini cannot relate to the person fighting fear of man at work.”

The Generation Effect Proves the Science

Cognitive science has a name for why sequence matters. It's called the generation effect — and it is one of the most replicated findings in learning research.

A meta-analysis of 86 studies found an effect size of .40 — almost half a standard deviation — for information you generate yourself versus information you simply read. When you wrestle with a passage, make your own connections, form your own outline before consulting anyone else, the material encodes into memory at a fundamentally deeper level.

When AI generates the outline first, you skip the generation phase. You consume instead of create. And the science is clear: consumed information does not stick the way generated information does.

That study — roughly 1,000 Turkish high school students, published in PNAS — showed that students with unrestricted ChatGPT access looked 48% better during practice. But when the AI was taken away and they had to perform alone, they scored 17% worse than the control group. The AI didn't build their math skills. It replaced them.

The Sequence Applied to Sermon Prep

Here is what the sequence problem looks like in practice:

Same tool. Same pastor. Same passage. Completely different spiritual and cognitive outcome. The variable is not the technology. The variable is the sequence.

Why the Church Keeps Missing This

The church's AI debate has two camps: fear and surrender. One side says ban AI entirely. The other side says let it write everything. Both are wrong because both focus on the tool instead of the workflow.

Eric Stoddart's taxonomy in Studies in Christian Ethics provides the framework. Churches can relate to AI in three postures: abdication (theologically unacceptable), delegation (acceptable with oversight), and collaboration (the target). The sequence problem determines which posture you're actually in.

A pastor who opens AI before opening the text has abdicated the study — even if he edits the output extensively afterward. A pastor who completes his study first and then delegates research verification to AI has achieved collaboration. The edit isn't what matters. The encounter is what matters.

The Pilot's Parallel

Aviation learned this lesson at the cost of lives. 77% of commercial pilots reported their flying skills had deteriorated from operating highly automated aircraft. When Air France Flight 447 lost its autopilot over the Atlantic, the stall warning sounded 75 times. The pilots couldn't recognize it. Air France's internal report cited a “generalized loss of common sense and general flying knowledge.”

The automation didn't fail. The pilots' ability to fly without automation failed. Because they had stopped practicing.

Bainbridge predicted this in 1983: the operator supposed to take over when automation fails is the one whose skills degraded from not practicing. That paper has 4,700+ citations. And it applies to sermon prep with uncomfortable precision.

The Practical Guardrail

The Monday rule is not legalism. It is wisdom. It is the same principle that tells an athlete to train before competing, a surgeon to study before operating, a pilot to fly manual regularly.

For pastors, we recommend a three-step sequence:

What OpenLumin Is Built For

OpenLumin was designed for step two. Not step one.

It is a research companion — not a sermon writer. It retrieves evidence from 15+ scholarly sources: commentaries, cross-references, original-language data, Ancient Near East cultural context, Theographic Bible data covering 3,000+ people, places, and events. Every claim is marked as verified or flagged for review.

It does not generate your outline. It does not write your application points. It does not replace your encounter with the text. It confirms what you've already found — and surfaces what you might have missed.

Because the question was never whether to use AI. The question is when.

About: AI Fluency Ministry helps the church understand and use AI wisely. OpenLumin is the practical application of that research — a free Bible research companion that retrieves evidence so you can do the thinking.