We don't predict.
We structure uncertainty.
A lab that publicly builds a track record. Every prediction tracked. Every miss published. Trust is earned, not claimed.
Every week: structured scenarios on signals that matter — with D/E/M/A uncertainty annotation, base-rate anchoring, and public Brier scoring. Free. Scored. Transparent.
Three layers of intellectual honesty
no one else provides
Where others give you opinions, we give you structured uncertainty — annotated, debated, and scored.
Multi-AI Debate
Multiple AI models independently analyze the same event, surface disagreements, and debate conclusions. The synthesis captures the strongest arguments from each side — not consensus, but structured dissent.
D/E/M/A Annotation
Every scenario is tagged with the D/E/M/A protocol: Evidence Level (L1–L4), confidence percentage, uncertainty type (epistemic, aleatoric, or model), black swans, and what event would critically change the assessment.
Public Retrotrack
Every prediction gets a resolution date and a Brier score. When predictions resolve, we publish what we got right, what we got wrong, and why. First scores due April 7-15, 2026. Calibration metrics will be published as data accumulates.
Six cross-cutting domains.
One unified intelligence feed.
Events don't respect boundaries between disciplines. Neither do we.
Geopolitical
Conflicts, diplomacy, regime changes, sanctions, trade policy, alliance shifts
Cyber & Threat Intel
APT activity, CVE attribution, infrastructure attacks, nation-state cyber operations
Markets & Macro
Geopolitical market impact, prediction market analysis, commodity disruption scenarios
Digital Sovereignty
AI regulation, data localization mandates, tech decoupling, compute export controls
Social Unrest
Mass protests, labor actions, political polarization, youth movements, anti-corporate campaigns
Climate & Disaster
Extreme weather, pandemics, resource scarcity, infrastructure failure, environmental crisis
8 lenses. One event.
See what others miss.
Every major scenario analyzed through local, regional, and global cultural lenses — sourced from local-language media, not Western assumptions about what other countries think.
Building trust by publishing
every prediction — and every miss
We just launched. Every prediction will be scored. Every miss analyzed. Calibration measured over time. This is transparency as a product — starting now.
We've published thirteen scenario analyses with 54 tracked predictions. Our first prediction has been scored: partial escalation on Iran was correct (Hit Rate 100%), but we were under-confident (Log Loss 0.916). Read the full retrotrack →
Free awareness. PRO depth.
Daily scenarios are free forever. PRO unlocks real-time capabilities and advanced analytical tools for professionals.
- 5 daily scenarios with D/E/M/A annotation
- Weekly retrotrack scorecard
- Friday deep dive analysis
- Scenario archive (searchable)
- Email digest (daily)
- Telegram channel access
- Vote for next week's scenario focus
- Real-time scenario alerts
- Personal calibration
- Test your own scenarios
Need enterprise-grade geopolitical simulation?
ScenarioAtlas.ai provides awareness scenarios. ScenarioAtlas.com provides full business impact simulation — modeling what happens to your specific company when you take a specific action during geopolitical events. InsightAtlas, DecisionAtlas, ImpactAtlas.