A fluent AI forecast can sound convincing. What evidence lets you inspect what was recorded, when, and how the outcome will be evaluated?
This episode separates forecast receipts, evaluation inputs and performance results. We explore correlated advisor perspectives, unresolved forecasts and why missing, inapplicable and zero values mean different things.
Dated audit: on October 4, 2026, a bounded metadata audit covered 746 public consensus records: 373 one-year and 373 five-year records. It checked six required metadata fields for presence. It did not recompute checksums, validate every field or establish forecasting accuracy.
Open the evidence: iPulse AI research · Historical consensus dataset · Open Forecast Receipt specification and verifier · Forecast Library
A matching digest can detect changed content. It cannot establish correct inputs, investment skill or publication before the event. Retrospective records need a retrospective label and stronger timing claims need separate evidence.
AI Investment Edge: Stocks, Markets & Money, by iPulse AI and Future Edge Group. Research by Russlan Ramdowar; conversation and voices generated with Google NotebookLM / Gemini Notebook. Research, not investment advice.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit