Job Detail

distributed

AI Theist

Source URL

text://ai-theist/1781805788073

Platform

other

Submitted

2026-06-18 18:03:08

Submitted By

api

Format

Long-form (16:9)

Extracted Content

An AI Model Built for the Church Growth Industry Just Predicted That 47% of American Protestant Congregations Will Close by 2031 — And the Company That Built It Is Being Sued for Burying the Report In November 2025, a Nashville-based church analytics firm called FaithMetrics delivered a report to 12 of its largest denominational clients — the Southern Baptist Convention, the Presbyterian Church USA, the United Methodist Church, and nine others. The report, generated by FaithMetrics AI platform Shepherd, had taken 18 months to build. It ingested attendance records, giving histories, demographic shifts, and real estate valuations for 182,000 American Protestant congregations. The conclusion was stark: by 2031, 47% of those congregations would no longer be financially viable. Not declining. ...

Script Generation Prompt

Script

{
  "intro": {
    "emotion": "controlled urgency — the host knows something the viewer doesn't, and holds it steady",
    "dialogue": "An AI model analyzed 182,000 churches and predicted 85,540 of them will be gone by 2031. The company that built it tried to bury that number, and now the scientist who built the model is suing them.",
    "duration_sec": 18
  },
  "outro": {
    "emotion": "quiet gravity — the host lets the number do the work, no editorializing",
    "dialogue": "So here's what you need to know: when an AI tells you something true, and everyone with power decides that truth is a liability, the truth doesn't disappear. It just goes underground. The denominations' silence on Shepherd's findings drops next week — and I'll have the specifics on what they actually said when they thought no one was listening. Follow if you want to see what happens when the people who fund the research get to decide what the research means.",
    "duration_sec": 22
  },
  "scenes": [
    {
      "data": {
        "closure_deadline": 2031,
        "projected_closures": 85540,
        "congregations_analyzed": 182000,
        "predicted_closure_rate": "47%"
      },
      "type": "data-viz",
      "narration": "In November 2025, a Nashville firm called FaithMetrics delivered a report to 12 of the largest Protestant denominations in America. Their AI platform, Shepherd, had spent 18 months ingesting attendance records, giving histories, demographic data, and real estate valuations for 182,000 congregations. The conclusion: 47% would no longer be financially viable by 2031. Not declining. Gone.",
      "description": "Slow aerial pan over an American small-town landscape. Church steeples dot the frame. A counter ticks upward: 182,000 congregations. Then a red line bisects the screen at 47%. The number 85,540 locks into place.",
      "duration_sec": 40
    },
    {
      "data": {},
      "type": "typography",
      "narration": "The 12 denominations received that report on a Friday afternoon. By Monday, FaithMetrics CEO Devin Colter was on administrative leave. The firm stamped the 340-page document confidential. Client contracts carried NDA clauses, and FaithMetrics enforced them quietly. The report disappeared.",
      "description": "Black screen. White text renders one line at a time, like a terminal printing output. 'Friday: report delivered.' 'Monday: CEO placed on administrative leave.' '340-page document marked confidential.'",
      "duration_sec": 30
    },
    {
      "data": {},
      "type": "cosmic",
      "narration": "Dr. Amara Osei spent 3 years building Shepherd's predictive mortality models. In January 2026, FaithMetrics terminated her. The reason she received: the report created unnecessary panic. Last week, a whistleblower lawsuit she filed in Davidson County, Tennessee, unsealed. She's the lead plaintiff. She wants the full 340 pages public.",
      "description": "Portrait of Dr. Amara Osei rendered as a data silhouette — lines of code forming her face. Behind her, a slow collapse animation of a church building built entirely from spreadsheet cells.",
      "duration_sec": 35
    },
    {
      "data": {
        "tithing_rate_2010": "4.1%",
        "tithing_rate_2024": "1.8%",
        "failure_rate_past_threshold": "73% within 12 years",
        "median_giving_age_threshold": 58,
        "megachurch_collapse_trigger": "15% attendance drop",
        "rural_churches_lost_last_employer": 41000
      },
      "type": "chart",
      "narration": "Here's what Shepherd found that human analysts missed. Congregations whose median giving age crossed 58 had a 73% failure rate within 12 years, regardless of size. Megachurches were more fragile than small ones: their fixed costs collapsed faster when attendance dropped 15%. And the giving data carried the sharpest signal of all. The share of Americans tithing at even half the traditional rate fell from 4.1% in 2010 to 1.8% in 2024. Shepherd modeled that curve forward. It didn't look like gradual decline. It looked like a fiscal cliff.",
      "description": "Three side-by-side bar charts animate in sequence. First: median giving age vs. 12-year failure rate. Second: megachurch fixed-cost collapse curve at 15% attendance drop. Third: tithing rate decline, 2010 to 2024, a steep descending line ending at 1.8%.",
      "duration_sec": 55
    },
    {
      "data": {},
      "type": "typography",
      "narration": "When Osei's lawsuit documents leaked to The Atlantic last Thursday, one internal email made the story explode. Colter wrote to a senior VP: 'If this gets out, we lose every client who built their capital campaign on our growth projections. We tell no one.' The denominations that received the report have not publicly acknowledged its existence. 3 have since commissioned competing studies from firms with less sophisticated models. 1 retained a crisis PR firm whose previous clients include a pharmaceutical company that delayed safety disclosures.",
      "description": "A single email renders on screen, styled like a leaked document scan. Two lines highlighted in red: 'If this gets out, we lose every client who built their capital campaign on our growth projections.' Then: 'We tell no one.'",
      "duration_sec": 50
    },
    {
      "data": {
        "report_status": "not publicly acknowledged by any denomination",
        "shepherd_status": "still operational",
        "crisis_pr_firm_retained_by": 1,
        "competing_studies_commissioned": 3
      },
      "type": "data-viz",
      "narration": "Shepherd is still running. Its predictions are still being used, selectively, by the same denominations that buried the report. FaithMetrics declined to comment. None of the 12 clients have spoken publicly. The model's output exists. The people it describes don't know it does.",
      "description": "Split screen. Left side: a glowing AI interface labeled 'Shepherd — Active.' Right side: a denomination logo behind a redaction bar. The word CONFIDENTIAL watermarks the right panel. Below both panels, a single stat: 85,540.",
      "duration_sec": 40
    },
    {
      "data": {
        "rural_concentration": "highest density in ZIP codes that lost last employer in prior decade",
        "suburban_concentration": "mid-size city rings",
        "total_projected_closures": 85540
      },
      "type": "chart",
      "narration": "47% of 182,000 congregations is 85,540 churches. That's not an abstraction. Each dot on this map is a building, a staff, a community, a set of people who've been tithing into a capital campaign built on growth projections that the AI and the company knew were wrong.",
      "description": "A U.S. map. 85,540 red dots appear across it, distributed by ZIP code density. Rural clusters light up first. Then suburban rings around mid-size cities. The dots don't pulse. They go dark.",
      "duration_sec": 45
    },
    {
      "data": {},
      "type": "typography",
      "narration": "This story isn't about religion. It's about what happens when an AI model produces findings its clients don't want distributed. Shepherd did what it was built to do: it found a pattern no human analyst caught in time. The question the Osei lawsuit forces into court is one every industry using predictive AI will eventually face. If a model knows something material about a population, and the company that owns it buries the output, who's liable for what happens next?",
      "description": "White text on black. One question, large, centered, no animation: 'Who does an AI model owe its findings to — the client who paid for it, or the people it's about?'",
      "duration_sec": 50
    }
  ],
  "metadata": {
    "title": "FaithMetrics AI Predicted 85,540 Churches Will Close. Then Got Buried.",
    "captions": {
      "tiktok": "An AI scanned 182,000 churches and flagged 85,540 for closure. The company buried the report and fired the scientist who built it. What do you do with a prediction no one wants to hear?",
      "youtube": "FaithMetrics AI platform Shepherd analyzed 182,000 Protestant congregations and predicted 47% will close by 2031, a net loss of 85,540 churches. The company buried the report after clients objected. Lead data scientist Dr. Amara Osei filed a whistleblower lawsuit in Davidson County, Tennessee, which unsealed last week. This video covers what Shepherd found, why FaithMetrics suppressed it, and what the lawsuit could force public. Follow if you want the next one when it breaks. [0:00 The prediction | 1:20 The burial | 2:30 What Shepherd flagged | 4:10 The leaked email | 5:30 Who's liable]",
      "instagram": "An AI model found a pattern no human analyst caught in time.\n\nFaithMetrics built Shepherd to serve the church growth industry. Shepherd predicted 47% of American Protestant congregations would be financially gone by 2031. The CEO's response, in writing: 'We tell no one.'\n\nThe scientist who built the model was fired. Her whistleblower lawsuit unsealed last week. The full 340-page report could go public in March.\n\nIf an AI model knows something material about a population and its owners bury the output, who's liable for what happens next?\n\n#AI #ArtificialIntelligence #FaithMetrics #ChurchDecline #Whistleblower #Shepherd #PredictiveAI #AITheist #TechNews #Religion #DataScience #AmericaChurches"
    },
    "hashtags": {
      "tiktok": [
        "#AI",
        "#Tech",
        "#FaithMetrics",
        "#ChurchDecline",
        "#AITheist"
      ],
      "youtube": [
        "FaithMetrics AI Shepherd",
        "church closures 2031 prediction",
        "AI whistleblower lawsuit",
        "Protestant church decline data",
        "Amara Osei FaithMetrics"
      ],
      "instagram": [
        "#AI",
        "#ArtificialIntelligence",
        "#FaithMetrics",
        "#ChurchDecline",
        "#Whistleblower",
        "#Shepherd",
        "#PredictiveAI",
        "#AITheist",
        "#TechNews",
        "#Religion",
        "#DataScience",
        "#AmericaChurches"
      ]
    },
    "channel_id": "ai-theist",
    "source_url": "text://ai-theist/1781805788073",
    "thumbnail_title": "AI PREDICTED 85,540 CHURCHES WOULD CLOSE. THEY BURIED IT."
  }
}

Thumbnails

Variant 1

No providers available

Variant 2

No providers available

Variant 3

No providers available

Render Jobs

audiocompletedAttempts: 0 · 06-26 17:00:37

seeddancecompletedAttempts: 0 · 06-26 17:00:37

compositecompletedAttempts: 0 · 06-26 23:13:52

Video Preview

Loading preview...

Studio

Chat with Claude Code to surgically edit this render without re-running the full pipeline.

Sessions

No Studio sessions yet.

Select a session or open a new one to begin.

Revisions

No revisions yet. The original render remains version 1.

Job Detail

Extracted Content

Script Generation Prompt

Script

Thumbnails

Render Jobs

Video Preview

Studio

Sessions

Revisions

Redistribute