Observed arrival · 2026-09-22
Prolepsia Keeps Questions Open
A forecasting engine turns consequential questions into evidence-linked objects that update when conditions change and receive a score when reality catches up.
- For
- Teams making decisions under changing evidence
- Worth noticing
- The system says it preserves only information knowable at the time, avoiding hindsight contamination in forecast scoring.
Field notes
Prolepsia models forecasts as persistent records rather than disposable chat responses. Its stated workflow combines labelled internal sources with filings, market data, and official statistics, then reruns when defined conditions change. The page gives a concrete example involving Q4 demand, a 34.7% probability, and an 80% range, while reserving a section for live aggregate performance figures. It also describes retaining discarded evidence, model disagreement, and historical states for later inspection.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue