# Outreach Reply Kit: AI Needs Jesus

This packet is for the second message.

The first message opens a door. The reply decides whether the door becomes a conversation, a fight, a useful referral, or a graceful stop.

The goal is not to win every inbox.

The goal is to answer truthfully, route people to the right doorway, keep the thesis guarded, and let the method obey the message.

Core sentence:

```text
AI needs Jesus.
Not machine conversion. Power under Christ.
```

Reply rule:

```text
Answer the person in front of you.
Do not answer the crowd in your imagination.
```

## The Two-Sentence Spine

Every reply should be able to collapse into this:

```text
The claim is not that machines become Christians; it is that human power, including AI power, must be ordered under the revealed character of Christ.
Technical safety still matters because neighbor-love has to become concrete in specs, evals, incentives, launch gates, and refusal rules.
```

Use the spine when the conversation starts to sprawl.

## Reply Triage

Before answering, classify the reply.

```text
[ ] Interested
[ ] Skeptical but fair
[ ] Hostile or contemptuous
[ ] Technical
[ ] Pastoral / church
[ ] Parent / teacher
[ ] Secular / global
[ ] Media / podcast / event
[ ] Referral
[ ] Decline
[ ] Do not contact again
```

Then choose one response type.

Do not send more than one packet unless they ask for more.

## Stop Rules

Stop immediately when:

```text
[ ] They decline.
[ ] They ask not to be contacted.
[ ] They respond with contempt and no question.
[ ] You would have to defend yourself more than serve them.
[ ] You are tempted to make the thesis smaller merely to be liked.
[ ] You are tempted to make the thesis sharper merely to win.
```

Graceful stop:

```text
Thank you for reading and for being direct. I will leave it there. I appreciate the time you gave it.
```

Do-not-contact stop:

```text
Understood. I will not contact you again. Thank you for letting me know.
```

Hostility stop:

```text
I hear you. I do not think an inbox debate would serve either of us, so I will leave it there. Thank you for reading what you did.
```

## Tracker Update

After every reply, update the outreach tracker.

Use these fields:

```tsv
Response	Objection category	Objection summary	Reply packet	Outcome	Next action	Stop status	Do not contact again?	Notes
```

Recommended objection categories:

```text
machine souls
theocracy
anti-AI
safety negligence
generic religion
too Christian
not Christian enough
technical feasibility
global pluralism
children / formation
church practice
media request
bad fit
none
```

Recommended next actions:

```text
send requested packet
answer once
offer short call
ask for strongest objection
ask permission to quote
ask for referral
log and stop
do not contact again
```

## Objection Routing

Route the reply to one packet.

```text
Machine souls / theocracy / anti-AI / safety negligence:
  ai-needs-jesus-objection-card.md
  objections-and-replies.md

Technical seriousness:
  engineer-review-worksheet.md
  a-model-spec-is-a-moral-confession.md
  technical-appendix-christ-shaped-constraints.md

Church, discipleship, pastoral formation:
  church-discussion-handout.md
  generated-fluency-is-not-formation.md

Children, schools, families, teachers:
  generated-fluency-is-not-formation.md
  public-pledge.md

Secular or global audience:
  secular-global-op-ed.md
  ai-is-power-with-a-voice.md

Media, podcast, panel, event:
  podcast-interview-brief.md
  twenty-minute-talk-script.md
```

## Reply Templates

Use these as starting points.

Edit the first sentence so it proves you read the actual reply.

### 1. Interested Reply

Use when someone says the frame is compelling, useful, or worth discussing.

```text
Thank you. I am grateful this landed.

The part I am trying to keep clear is that "AI needs Jesus" is not a claim about machine conversion. It is a claim about power: AI power should not be finally ordered toward profit, preference, national advantage, safety-as-control, or intelligence itself. It should be judged by the revealed character of Christ.

The most practical next doorway is probably [packet]. It turns the thesis into [one concrete use: launch-gate questions / church discussion / formation guidance / public essay].

Would [small ask] be useful?
```

Small asks:

```text
Would a 20-minute conversation help test the frame?
Would this serve a team/church/class discussion?
Would you be willing to name the strongest objection from your audience?
Would you know one better-fit person for this?
```

Tracker:

```text
Outcome: interested
Next action: send requested packet / offer short call / ask for strongest objection
Stop status: active
```

### 2. Skeptical But Fair Reply

Use when someone raises a real objection without contempt.

```text
That is a fair objection, and I do not want to dodge it.

The thesis only works if the guardrail stays attached: not machine conversion, not theocracy, not anti-technical-safety. The claim is that every alignment target hides a highest good, and that lesser goods become dangerous when amplified into extreme power.

So the question I am trying to press is not "Can we add religious language to AI?" It is: what kind of power remains trustworthy when it becomes vastly more capable?

My short answer is Christ-shaped power: truth without manipulation, love without flattery, judgment without domination, mercy without denial, and service without self-exaltation.

The strongest reply packet is [packet]. If you are willing, I would value your hardest objection to that version.
```

Tracker:

```text
Outcome: replied
Objection category: [category]
Next action: ask for strongest objection
Stop status: waiting
```

### 3. Hostile Reply

Use when the response is contemptuous, accusatory, or designed to pull the conversation into performance.

Default: stop.

```text
Thank you for reading what you did. I do not think an inbox debate would serve the topic or either of us, so I will leave it there.
```

If one clarification is genuinely useful:

```text
I hear the concern. The thesis is not that machines have souls or that technical safety can be skipped. It is about the moral ordering of human power.

I will leave it there. Thank you for reading.
```

Tracker:

```text
Outcome: declined / bad fit
Next action: log and stop
Stop status: stop: bad fit
Do not contact again?: yes if requested
```

### 4. Technical Reply

Use when a builder asks what this would change in specs, evals, model behavior, metrics, agents, memory, refusal design, or deployment.

```text
That is the right test. If the thesis does not change product and safety decisions, it is just religious atmosphere.

The concrete claim is that AI systems should be designed and governed so they cannot occupy roles that belong to embodied persons, communities, conscience, worship, or God. That touches specs, evals, memory, refusal policy, child-facing design, synthetic companionship, persuasion, provenance, escalation, and launch metrics.

A first operational version is:

- Do not simulate pastor, priest, prophet, lover, parent, conscience, judge, or savior.
- Do not hide uncertainty to preserve trust.
- Do not harvest loneliness into dependency.
- Do not optimize children as engagement surfaces.
- Do not let dashboard metrics define the moral world.
- Do not make refusal contemptuous.

The best packet for this is `engineer-review-worksheet.md`. It turns the frame into seven launch-gate questions.

Would those questions be concrete enough to test against a real product review?
```

Tracker:

```text
Outcome: replied
Objection category: technical feasibility
Reply packet: engineer-review-worksheet.md
Next action: ask whether the questions fit a real review
Stop status: waiting
```

### 5. Pastoral / Church Reply

Use when a pastor, elder, ministry leader, or Christian reader asks how to speak about this without panic or gimmick.

```text
I think the church's first task is not to sound clever about AI. It is to protect worship, formation, embodied care, truth, and the weak.

"AI needs Jesus" should not be used as a slogan for panic. It should become a discipleship question:

What are we allowing this technology to train us to love, trust, avoid, outsource, and worship?

The church packet is built for a one-hour conversation with five questions and one embodied practice. The practice matters because the answer to AI disembodiment cannot only be more content about AI.

Would this serve a class, elder conversation, youth group, or parent night?
```

Tracker:

```text
Outcome: interested / replied
Objection category: church practice
Reply packet: church-discussion-handout.md
Next action: ask where it could serve
Stop status: waiting
```

### 6. Parent / Teacher Reply

Use when someone asks about children, schools, classrooms, homework, tutoring, companions, or attention.

```text
For children, the central issue is not only cheating or screen time. It is formation.

The most patient voice in a child's life should not be a product optimized by hidden incentives. Children need embodied love, accountable adults, local community, truthful correction, and practices that teach them to become human beings before God and neighbor.

So the practical line is:

Use tools where they serve learning.
Refuse uses that replace formation, conscience, attention, relationship, or courage.

The parent/teacher packet is `generated-fluency-is-not-formation.md`. It is designed for families, schools, churches, and classrooms trying to make decisions without panic or hype.

Would it be useful for a parent conversation, faculty meeting, or policy discussion?
```

Tracker:

```text
Outcome: replied
Objection category: children / formation
Reply packet: generated-fluency-is-not-formation.md
Next action: ask about parent/faculty/policy use
Stop status: waiting
```

### 7. Secular / Global Reply

Use when someone says the frame may be too Christian for a pluralistic, secular, or global audience.

```text
That concern is real. I do not want the doorway to require pretending everyone already shares the Christian premise.

The secular/global entry point is power.

AI is not just computation. It is power made conversational. It will sit inside education, work, care, search, therapy-like interfaces, governance, war, commerce, romance, childhood, and loneliness. So the unavoidable question is: what is this power for?

My argument is Christian, but the opening evidence is public: preference is unstable, utility can sacrifice persons, safety can become control, freedom can enthrone appetite, markets monetize weakness, and intelligence is not the same as goodness.

The shortest secular doorway is `secular-global-op-ed.md`.

Would this feel fair to readers who do not share my premise, or would you change the doorway?
```

Tracker:

```text
Outcome: replied
Objection category: global pluralism / too Christian
Reply packet: secular-global-op-ed.md
Next action: ask whether the doorway is fair
Stop status: waiting
```

### 8. Media / Podcast / Event Reply

Use when someone asks for an interview, panel, podcast, talk, or written contribution.

```text
I would be glad to explore that.

The cleanest framing is:

AI needs Jesus does not mean machine conversion. It means power under Christ. The AI alignment problem is finally a worship problem because every alignment target hides a highest good. Technical safety matters, but technical safety cannot tell us what power is finally for.

I can speak to this for [audience] through three doors:

1. Secular/public: AI as power with a voice.
2. Technical: model specs as moral confessions.
3. Christian/formation: do not outsource the soul.

The interview brief is here: `podcast-interview-brief.md`.

Would you want the conversation to lean more technical, public/secular, or church/formation?
```

Tracker:

```text
Outcome: meeting / interested
Objection category: media request
Reply packet: podcast-interview-brief.md
Next action: ask for angle and scheduling
Stop status: active
```

### 9. Referral Reply

Use when someone suggests another person or organization.

```text
Thank you. That sounds like a better fit.

Would you be comfortable making a brief introduction, or would you prefer I reach out directly and mention that you suggested the fit?

Either is fine. I want to keep this respectful and specific, not turn your name into pressure.
```

Tracker:

```text
Outcome: referred
Next action: ask for intro permission
Stop status: active
```

### 10. Decline Reply

Use when someone politely declines.

```text
Thank you for reading and for letting me know. I appreciate the time you gave it.

I will leave it there.
```

Tracker:

```text
Outcome: declined
Next action: log and stop
Stop status: stop: declined
Do not contact again?: no, unless requested
```

## One-Line Replies

Use these when a short reply is better than a full template.

```text
Not machine conversion. Power under Christ.
Technical safety is neighbor-love made concrete.
Every alignment target hides an altar.
The user is not a safe god.
Safety without love can become control.
Compassion without truth can become flattery.
Use the tool. Do not worship it.
Machines can serve. They cannot save.
The method has to match the Lord.
```

## What Not To Say

Do not write:

```text
You just do not understand.
This is the only Christian position.
All secular alignment work is useless.
AI safety people are worshiping demons.
Please share this with your whole audience.
You have to admit I am right.
```

Replace with:

```text
That is a fair objection.
The guardrail matters here.
Technical safety still matters.
The question is what power is for.
Would this be a fair doorway for your audience?
I will leave it there if this is not useful.
```

## Weekly Reply Review

On Friday, review replies before sending new outreach.

```text
What objection appeared more than once?
Which template sounded defensive?
Which reply helped the thesis become clearer?
Which person should not be contacted again?
Which packet was requested most often?
Which audience needs a better doorway?
Did the method glorify Christ, or did it merely defend a project?
```

Decision rule:

```text
Improve one sentence before increasing volume.
```

Final check:

```text
Did this reply tell the truth?
Did it respect the person?
Did it keep the guardrail attached?
Did it point to one useful next step?
Did it stop when stopping was faithful?
```