Three Maps to 2038

AI Scenarios to Sharpen Civil Society's Strategic Capabilities

Three futures. Three demands. One sector that cannot afford to guess wrong. This report stress-tests Civil Society against stagnation, progress, and rupture — and finds that in every future, the world needs more from us, not less.

The window is closing. Path dependencies are forming. Somewhere between the model that hallucinates and the one that doesn't, between the climate deadline and the governance vacuum, the foundation of the next fifty years is being poured. Civil Society has not been in the room. It needs to be.

Built on the Oxford Scenario Planning Approach and grounded in narrative futures set in England, Kenya, and Singapore, this is the strategic case for building new capabilities now — before the foundation sets.

Read the full report

Who is this for?

five doors into the same building

The Interdependent System
community intelligence regulatory frameworks adaptive capacity infrastructure funding external diagnostic technical investment living data theoretical frameworks co-investment · independent evaluation regulation · compliance evidence base · evaluation mandates responsible deployment · partnership research funding · evidence access · independent assessment · frontier questions Civil Society Leaders the expanded mandate Policymakers the verification layer Funders & Foundations the infrastructure gap AI Labs & Tech Companies the mirror you need Researchers & Social Scientists the outgrown frame Civil Society is the most connected node — not because it governs the others, but because it is the only one accountable to the people experiencing consequences. Civil Society flows Peer-to-peer flows The Promethean Collective · theproco.org

The people in Three Maps.

The Thinking Behind the Work

Companion essays

What If We're Wrong? Three null hypotheses. The intellectual honesty this work required.

What If We Win? What Civil Society actually looks like when it rises to the moment.

Holding Scenario A Lightly Exploring small but consequential variations in the baseline stress test.

How These Were Made: A note on method, detailing our LLM-supported twist on the Oxford Scenario Planning approach.

How These Were Made An AI-assisted approach to Oxford Scenario Planning STAGE ONE Structured Multi-Model Research Ensemble 5 models · 10 PESTLE-assigned specialist instances · shared research question ① PARALLEL DRAFT — independent, no cross-visibility CLAUDE Political Anthropic CLAUDE Economic Anthropic CHATGPT Social OpenAI CHATGPT Technological OpenAI GEMINI Legal Google GEMINI Environmental Google GROK Political xAI GROK Economic xAI PERPLEXITY Social Research PERPLEXITY Technological Research instance 1 instance 2 ② MUTUAL PEER REVIEW — each assesses all 9 others: methods · analysis · conclusions · blind spots ③ REVISED SECOND DRAFTS — each instance incorporates critique from all 9 peers SYNTHESIS ACROSS 10 SECOND DRAFTS convergence · divergence · genuine novelty DRAFT SCENARIOS STAGE TWO Prism Council 18 analytical personas · philosophy · ethics · strategy · critical theory ARCHITECTS 6 lenses Strategy Systems Governance Economics Technology History Meadows · Rumelt HEALERS 6 lenses Ethics Care Postcolonial Justice Psychology Community Lorde · Foot · Srinivasan VISIONARIES 6 lenses Philosophy Critical Theory Futures Culture Religion Art Arendt · Taylor Each scenario submitted to all 18 for independent stress-testing of assumptions, consistency, blind spots Human judgment governed synthesis and all editorial decisions throughout OUTPUT Three Maps to 2038 Scenarios · Executive Briefings · Companion Essays · Visual Package + APPENDIX A: FULL METHODOLOGY "We are transparent about the role of AI in producing this work, because transparency about tools is part of what we are asking of the sector."

Be First to Shape What Comes Next

We are building the first Civil Society AI curriculum. Not a tool guide. Not a policy primer. A genuine capacity-building programme for organisations that want to lead.

Sign up to hear when it's ready — and to help shape it.

[We will not flood your inbox. When we have something worth saying, we will say it.