- Purpose
- An evidence-literacy game in which teams interrogate conflicting dossiers, assemble limited specialists and commit one coherent operation.
- Format
- Team table + dossiers + approved AI + reveal
- Decision arc
- Evidence → specialists → five-part plan → report
- Public demo
- 3–4 minutes
An evidence-literacy game in which teams interrogate conflicting dossiers, assemble limited specialists and commit one coherent operation.
What it explores
The Uninsurables examines evidence discipline, collaboration and planning when AI can help a team search but cannot decide what constitutes a coherent case. Conflicting documents contain plausible stories, missing facts and details that matter only in combination.
The challenge is not to produce the most confident explanation. It is to build a plan that remains consistent across evidence, language, custody, route, ethos and available assets.
How the experience works
Teams work with a field manual, PDF dossiers and an approved AI tool. They select one of ten operations, allocate three limited-use specialists and submit five linked decisions. Credentials, heat and losses persist between operations.
The server adjudicates deterministically. Individually attractive choices can fail when they contradict one another, allowing the debrief to focus on reasoning rather than chance.
The consequential moment
The after-action report identifies where the team’s plan stopped being coherent and points back to the evidence that was missed or overinterpreted. The useful output is a shared account of how the story became more persuasive than the facts.
Public demonstration
A compressed loop uses three short evidence fragments, two possible routes and one specialist decision. The reveal explains coherence and missed evidence rather than awarding a generic intelligence score.