Lecture 15-3: Part III Synthesis — The Boundary Between Tool and Ministry

1 Corinthians 4:1-2 — "So let a man think of us as Christ's servants, and stewards of God's mysteries. Here, moreover, it is required of stewards, that they be found faithful." (World English Bible)

2 Corinthians 4:7 — "But we have this treasure in clay vessels, that the exceeding greatness of the power may be of God, and not from ourselves." (WEB)

3.1 The boundary, restated once

Stated plainly, without the qualifications each individual chapter needed to make it fair: a task is a candidate for AI assistance to the degree that it is informational (retrieving or organizing existing knowledge), logistical (scheduling, routing, administering), or preparatory (drafting, researching, first-pass work reviewed before it reaches anyone). A task is not a candidate for AI substitution to the degree that it is relational (it depends on a particular person knowing and being known by another particular person), pastoral (it carries responsibility for someone's spiritual and often practical wellbeing), or sacramental (it is constituted by an authorized person's action in a specific, embodied setting, not merely by the correct words being said).

This is a better test than "low-stakes versus high-stakes," which the course has deliberately avoided, because some informational tasks are extremely high-stakes (a policy affecting how sensitive pastoral data is handled, chapter 15's own subject) and some relational tasks can look, from the outside, almost administrative (a hospital visit that is mostly sitting in silence, as Job 2:13 modeled in 12-2). Stakes measure how much is riding on getting something right; the informational/relational distinction measures what kind of thing is actually being done, and only the second question determines whether AI assistance changes what is happening or merely helps someone do it.

3.2 The pattern across five chapters

Chapter 11, preaching. AI is a legitimate research aid — word studies, cross-reference lookup, structural brainstorming (11-1) — and never a substitute for the preacher's own wrestling with the text (Genesis 32:24's Jacob, alone through the night) or for the Spirit's work of illumination in the person who will stand and preach (11-2's distinction between inspiration, illumination, and information retrieval). The boundary sits exactly at authorship and conviction: research assisted, content and conviction not delegated (11-3).

Chapter 12, counseling. AI chatbot support is genuinely useful for some of what people actually seek from it — availability, low perceived judgment, some documented short-term relief (12-1) — and it cannot supply incarnational presence, the shared vulnerability Hebrews 4:15 grounds in Christ's own temptation "in all points as we are," or decisive human intervention in a crisis, where Acts 16:28's jailer scene sets the standard of what is actually required (12-3). The boundary sits at presence and intervention: information and availability assisted, embodied co-suffering and crisis response never delegated.

Chapter 13, fellowship. Livestreaming and digital tools can genuinely extend pastoral reach to the homebound and geographically isolated (13-2), and they cannot constitute koinonia in the full, embodied sense the New Testament actually describes — the "breaking of bread" of Acts 2:42, the "one bread, one body" of 1 Corinthians 10:17 — nor, in the view held by most Christian traditions surveyed in 13-3, validly carry the Lord's Supper absent physical presence and an authorized celebrant. The boundary sits at embodiment and sacrament: reach extended, shared physical presence not replaced.

Chapter 14, discipleship. AI tools can genuinely help with Bible study logistics — cross-references, word studies, generating discussion questions (14-2) — and cannot form a disciple, because formation (Galatians 4:19's "formed in you") is a slow, relational, imitative process (1 Corinthians 11:1's "imitate me") that information transfer, however efficient, does not touch. The boundary sits at formation itself: information assisted, the slow relational shaping of character never delegated.

Chapter 15, administration. AI tools can genuinely assist scheduling, routine drafting, and first-pass transcription (15-2's permitted-uses list), and they cannot be handed accountability for decisions about a specific member's spiritual condition or given custody of a congregation's most sensitive disclosures without a named, accountable person governing exactly how (15-1's reading of Acts 6; 15-2's data-handling standards). The boundary sits at accountability itself: logistics assisted, stewardship of trust never delegated.

The pattern is the same pattern each time, appearing in five different vocabularies because five different ministry areas have five different technical vocabularies. Naming it once, as this lecture does, is meant to make the next new AI tool or feature — one this course has not anticipated, because none of the four preceding lectures wrote a rule that mentions it by name — easier to place correctly the first time, by asking which side of the boundary the specific task actually falls on, rather than treating each new tool as an unprecedented question.

3.3 Stewardship as the governing image

1 Corinthians 4:1-2 supplies the frame that holds chapters 11 through 15 together as one argument rather than five separate rulings. A steward, in the first-century household sense the text assumes, manages resources that belong to someone else, exercises real discretion in doing so, and remains personally accountable for how that discretion is used. This is a precise description of the church's actual relationship to AI tools across every area this course has examined: the tools are genuinely useful (real discretion is being exercised in choosing to use them), they are not the church's own to use however is most convenient (they touch what belongs to God — the gospel entrusted to be preached, the members entrusted to be shepherded, the sacrament entrusted to be rightly administered, the data entrusted in confidence), and "found faithful" (4:2) is a standard of accountability, not merely of good intentions or technical competence. The steward image is why chapter 15's administrative concerns are not a lesser, merely practical coda to the more obviously theological chapters 11 through 14 — administration is exactly where the stewardship of what has been entrusted is tested in its most mundane, and therefore most easily neglected, form.

3.4 The vessel was never the source of the power

2 Corinthians 4:7's "treasure in clay vessels" closes Part III with a claim that cuts in both directions and should not be softened into only one of them. It is, first, a caution against overestimating what an impressive AI tool can contribute to ministry: the power was never in the vessel's sophistication, so a more sophisticated vessel does not thereby carry more of the treasure. But it is equally, and just as importantly, a comfort against underestimating an unsophisticated ministry: a small congregation with no advanced software, an under-resourced pastor with no research assistant of any kind, has not thereby been handicapped in the one thing 2 Corinthians 4:7 says actually matters, because the treasure's power was always "of God, and not from ourselves" — a claim equally true of a technologically bare ministry and a technologically well-equipped one. Part III's cumulative argument is not that technology is dangerous and should be minimized out of caution; it is that technology, wherever it appears in ministry, is simply not the kind of thing that could ever have been the source of the power in the first place, which is exactly why the boundary this lecture has restated matters — not because AI is uniquely threatening, but because confusing the vessel for the treasure was always a mistake, with or without AI to make it newly tempting.