Observation is not prediction
What is measured today can inform but not guarantee tomorrow. We label verified observations, hypotheses, conditional scenarios and scored forecasts differently.
Futures Observatory.
Our forecasts are dated propositions with explicit uncertainty and possible disconfirming evidence.
EVIDENCE FRAMEWORKREVISION-READYOPEN QUESTIONS
WHAT THIS FIELD EXAMINES
Scenarios, indicators and claims that can be tested over time. Our purpose is to establish the evidence and questions needed for rigorous coverage, not to imply that a proposed future is inevitable.
Pre-register before scoring
A testable forecast needs a resolution criterion, deadline, source, baseline probability and update history. We do not backfill certainty once an event occurs.
An open-question register
We begin with research questions—not invented forecast accuracy statistics. The scoring register becomes operational when dated, measurable forecasts are actually published.
How we would test the claim.
Record forecast statements before resolution, preserve all revisions and evaluate calibration later.
OPEN QUESTION REGISTER / NOT A SCORED FORECAST
Which changes will be measurable?
Use these prompts as research questions. We will only call an entry a forecast after publishing a testable statement, due date, resolution source and timestamped probability.
- How does AI change completed-task time for matched developer groups?
- How much of a benchmark gain survives independent tests?
- When do tools reliably preserve audit trails and human control?
Documents to consult
These are starting references for future reporting, not proof that every open question above is resolved.
Revision record
Edition 1.0 · 9 October 2026. Initial scope and source register created. This brief is a research framework, not a completed investigation; material future corrections should be described rather than silently overwritten.
How corrections are documented ↗