# Project Imagination — Research question

Status: unresolved

## Current working formulation

How can deliberate possible-world construction, manipulation, exploration, and evaluation be defined and measured, and what would constitute reproducing that capability in an AI system?

This formulation follows the initialization brief. It does not establish a distinct cognitive mechanism or imply that current AI systems lack relevant abilities.

## Priority open questions

1. What makes an activity deliberate possible-world construction rather than prediction, simulation, planning, creativity, or innovation as studied in Project Fiction?
2. What aspects of a constructed world must be represented, and how would manipulation, exploration, and evaluation be observed?
3. How should consistency, responsiveness to changed assumptions, and evaluative behavior be assessed, if they are adopted as criteria?
4. Which tasks and comparisons distinguish the intended capability from a plausible hypothetical description or recall of familiar scenarios?
5. Which evidence would support a computational reproduction claim, and where do existing systems already meet possible criteria?

## Resolution status

The operational definition, task set, evaluation criteria, and research hypotheses are unresolved. Proposed criteria above are questions to examine, not an adopted benchmark. Update [methodology](methodology.md) and [research status](research-status.md) when a position is adopted, with its supporting rationale and history.