Every major current event has a historical echo. Déjà Vu Weekly finds the pattern, scores the strength of the analogy, and makes predictions we track publicly.
Each issue follows a fixed 10-section format: identify the flashpoint, find the historical parallel, measure the structural overlap, assign a pattern strength score, surface what people missed last time, model forward scenarios with explicit probabilities, and track every prediction we make.
No vibes. No spin. Just structural analysis with accountability built in.
Déjà Vu Weekly is produced with the help of Claude Code, an AI assistant by Anthropic. Each issue is researched, written, and published using Claude as a collaborative tool.
Here's what that means in practice:
We believe transparency about AI involvement is the right default. You deserve to know how what you're reading was made.
Anyone can explain the past. The prediction tracker is what keeps us honest. Every scenario we model gets a probability. Every probability gets graded against reality. Over time, the tracker reveals whether our historical analogies actually have predictive power — or whether we're just telling good stories.
Déjà Vu Weekly is created by Gregg Carey. Questions, corrections, or pattern suggestions are welcome at gregg.e.carey@gmail.com.