Observed arrival · 2026-10-28
Truth & Reconciliation’s daily AI forecast ledger
A research project compares daily AI-model forecasts on real event-market questions and scores them after settlement.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Readers evaluating AI forecasting claims
- Worth noticing
- The described ledger retains failures, timeouts, and refusals instead of dropping them from the forecast record.
Field notes
The described round selects open event-market questions, freezes rules, official data, and headlines at a fixed UTC cutoff, then records model answers with their inputs. Forecasts are not scored until the questions settle, using the Brier score; the page explains that an always-50% forecast scores 0.25. Each model has a news seat and a rules-only twin, while a seventh reference seat ignores the question.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue