Fiction Lab / Collection
AI evidence and stewardship
Move from evidence for AI authority to long-term ownership, migration and exit.
Collection route
- 1Research desk
Blindsight
What evidence should an organisation require before treating an apparently intelligent system as a competent decision-maker?
Challenge fluent competence.
Open - 2Theme
Intelligence and evidence
The evidence required before a capable-looking system receives decision authority.
Frame the evidence threshold.
Open - 3Research desk
The Lifecycle of Software Objects
Who remains responsible for an intelligent system after the original product, vendor or team has ended?
Extend responsibility across time.
Open - 4Case
ARGUS Claims
Set the next autonomy level using segmented evidence rather than fluent explanations or aggregate agreement.
Choose bounded authority.
Open - 5Case
NURTURE OS
Respond to vendor exit without losing state, hiding behavioural drift or preserving an unsafe legacy indefinitely.
Design a continuity path.
Open