CONTINGENCYPLAN.AISign In

Methodology

How It Works

A daily Exit Signal Score, built from historical pattern recognition, daily global signals, and the consensus of six independent AI models.

We studied what collapse actually looks like

30+ instances of state collapse over 250 years — from the French Revolution to modern Venezuela. We identified patterns, warning signs, and tipping points so we know what to watch for.

We monitor what's happening every day

Every day, we pull from global news sources and social media. Hundreds of data points about what's happening in the US — so you don't have to spend hours scrolling to piece it together yourself.

an ensemble of AI models score the situation independently

multiple frontier models each analyze the day's events and assign a score from 0-100. We average them and show you all six — because no single model should be trusted alone.

You get a 2-minute briefing every morning

Every morning you get the Exit Signal Score, a short assessment of what's going on, and the top 5 things you should know about. That's it. No doom scrolling required.

Get Started →

FREQUENTLY ASKED

Frequently asked questions

How is the Exit Signal Score calculated?

Every day, we generate a briefing from global news and social media signals. an ensemble of independent frontier AI models — multiple frontier models — each read that same briefing and score the US situation from 0 to 100. The consensus score is the arithmetic mean of the models. All all individual scores are published alongside the consensus so readers can see where the models agree and disagree.

What historical data is the methodology based on?

We studied 30+ instances of state collapse over the last 250 years, from the French Revolution to modern Venezuela. We extracted the patterns, warning signs, and tipping points that appeared consistently across those cases. Those patterns are the lens each model uses when reading the daily briefing — not a specific prediction, but a framework for recognizing unstable conditions.

Why use multiple AI models instead of one?

Every language model has its own training data, cutoffs, biases, and blind spots. Any single model can be manipulated, can hallucinate, or can simply be wrong. Running the same briefing through multiple models from different labs dramatically reduces the chance that idiosyncratic errors shape the score. When all six agree, the signal is strong. When they disagree, that disagreement is itself informative and we show it.

What data sources feed the daily briefing?

We pull from global news wires, structured data providers, and public social media signals. Events are clustered, deduplicated, and categorized before they reach the models. The goal is a representative briefing of the day — not a cherry-picked list of alarming headlines. Models score based only on the briefing, so the briefing is the thing that shapes the score.

Can the score predict when to leave?

No. The Exit Signal Score describes present conditions, not future events. It is a preparedness tool — it tells you how much attention the situation warrants today, so you can plan in stages instead of reacting under pressure. It does not forecast specific events, dates, or decisions. CONTINGENCYPLAN.AI does not give financial, legal, immigration, or political advice.

Who runs this site?

CONTINGENCYPLAN.AI is an independent preparedness platform. It is funded primarily by a $5/month supporter tier. The site is not affiliated with any political party, government agency, news outlet, or AI lab. Scoring is automated and the methodology is public — everything on this page.

CONTINGENCYPLAN.AI

AboutMethodologyPricingPrivacyTerms

© 2026 NSMBL.IO LLC. CONTINGENCYPLAN.AI is a publication of NSMBL.IO LLC.