A forecast claim tracker decomposes a broad scenario into testable statements about events, quantities, dates, capabilities, or economic conditions. Each claim retains its original text, interpretation, evidence, status, uncertainty, and update history.
Good tracking distinguishes confirmed, partly supported, unresolved, contradicted, and not-yet-due claims. It also prevents hindsight bias by preserving the initial forecast and documenting judgment calls instead of quietly adjusting milestones to fit later reality.
ELI5
A forecast claim tracker records specific predictions and later compares them with dated evidence. It preserves the original wording so people cannot quietly rewrite a forecast after seeing what happened.
For example, a broad AI scenario can be divided into claims about capability, dates, costs and infrastructure. Each claim can remain unresolved, become partly supported, be confirmed or be contradicted, with the evidence and judgment history kept beside it.
