Proposal · GIA
Generated Intelligent Autonomy
“How can intelligent systems adapt and coordinate meaningful actions within delegated authority?”
Five proposed assessment dimensions
- 1Delegated decision authority
- 2Adaptive planning and recovery
- 3Context and state continuity
- 4Workflow and resource coordination
- 5Regulatory capacity: monitoring and maintaining acceptable operating conditions
The homeostasis analogy
Living systems keep temperature, chemistry and energy within safe ranges. GIA proposes that useful autonomy works similarly: a system continuously monitors and maintains acceptable operating conditions rather than simply executing steps.
Capability is not permission. GIA scores what a system can do separately from what it is authorized to do. High capability with narrow delegation is a valid, often desirable, configuration.
Limitations and validation
Open question Any numerical scales or models are provisional and not yet validated.
Validation plans: compare against existing autonomy-level taxonomies, test inter-rater reliability, and pilot on documented agent workflows. Working papers will be published here as versioned drafts.
Challenge the model →