A theology of design
Most software is built without theological commitments. This one is not.
The 3RC Discernment Companion is a tool that operationalizes FaithTech's Redemptive Technology Playbook (faithtech.com/playbook) for leadership teams considering AI adoption. Every design decision in this app comes from a specific commitment to a specific theology. This page is the documentation of that commitment.
Why we begin with lament
The first thing this tool asks you to do is not frame your decision. It asks you to sit with the pain that made the decision necessary.
This is from Day 1 of the CHAI Prompting Bible, which itself draws on the prophetic tradition of Israel — Jeremiah, Lamentations, the Psalms of lament. Discernment that skips lament becomes optimization. Optimization is what got us here.
By naming who is hurting before naming what we'll build, the tool resists the technologist's default move of "let's solve this." Some problems should not be solved with technology. We need to feel that first.
Why 3RC has four postures, not just one
Most teams considering AI immediately ask: "What should we build?" The 3RC framework refuses this question for as long as possible. It asks instead: should anything be built at all?
Reject is first because the most faithful answer to most AI questions is no. Receive is second because if something must be done, someone else may have already done it more faithfully. Reimagine is third because adapting another's faithful work is humility. Create is last because it requires the most accountability and the most risk of harm.
The tool gates Create explicitly. You cannot reach Step 5 unless your team has genuinely engaged with Steps 2, 3, and 4 and produced honest negative verdicts. This is by design. Most AI projects in the church and the development sector skip straight to Create. This tool refuses.
Why we pause between steps
Between every step, this tool inserts a Selah — a contemplative pause. A verse, an inhale phrase, an exhale phrase, and twelve seconds you cannot skip past.
This is not a UX gimmick. It's a theological claim: that the rhythm of breath, scripture, and silence is itself part of discernment. Without it, you have an analytical exercise. With it, you have a practice.
The breath prayers are drawn from scripture passages that speak to the moment you're in. After Reject, be still and know. After Receive, I planted, Apollos watered, God gives growth. After Reimagine, have the same mindset as Christ Jesus. After Create, he made himself nothing, taking the form of a servant. Each one is a theological corrective for the specific temptation of that step.
You can turn them off. Some teams will move fast. We respect that. But the default is on, because the practice is the point.
Why we default to kenotic design
The Christian doctrine of kenosis comes from Philippians 2:5-11 — Christ emptied himself, took the form of a servant, became obedient even to death. Kenotic design is the discipline of building less. Of leaving room for human judgment. Of refusing to automate what should not be automated.
You will see this throughout the tool. The AI reflections suggest smaller versions. The Create step asks "what's the smallest version you could ship?" The recommendation often ends with "probably never." These are not bugs. They are the theology in working form.
Why the AI pushes back
The AI reflections in this tool are not a cheerleader. They are configured to be honest — to name what's strong AND what's rushed, vague, or self-serving. To affirm and to push back.
This is intentional. The MoReBench research from Scale AI (October 2025) found that current frontier AI models systematically underperform on Virtue Ethics, Contractualism, and Contractarianism — the moral traditions closest to most Christian thought. They favor Benthamite Utilitarianism and Kantian Deontology, which sound rigorous but often miss the human stakes.
We cannot fix this in the model. We can only configure the AI to be aware of its own partiality and to push the team to do the work the AI cannot. The system prompt you cannot see is designed for exactly this. The AI's job is to make your team more discerning, not to make your discernment easier.
Why we cite FaithTech and CHAI throughout
Receive is a posture, not just a step. This tool stands on FaithTech's framework, the CHAI Prompting Bible, the work of Mike Lee and the Scale AI moral reasoning team, the witness of the Christians in AI community.
Citing the source is itself a theological move. We are not the originators. We are participants in a longer tradition.
What this tool is and isn't
This tool is a prototype, built for the Christian Ethics of AI workshop at the University of Notre Dame on June 10, 2026. It is the operational form of a framework that has been forming for years.
It is not a replacement for your team's actual discernment. It is a scaffold for that discernment.
It is not theologically neutral. It cannot be. Neutrality in AI design is a lie — every default is a value, every absence is a choice, every interface is an argument. This tool argues, openly: that technology should love the people involved, that less is often more, that the human in the loop is the feature.
If you build redemptive AI for your context, we'd like to know. Tell us at chaiglobal.ai.
— Nathan Gaw and Richard Zhang, CHAI Global · in partnership with FaithTech