Framework
Human Capability Map
A research-in-progress attempt to inventory human capabilities that matter for civilization—and which are most exposed to automation.
Problem
Public debate about AI and education often treats 'human capability' as a slogan instead of a map with parts that can atrophy, transfer, or remain essential.
Map human capabilities across perception, modeling, craft, care, judgment, coordination, and meaning-making—then track exposure to automation versus need for human stewardship.
Model
Problem
Without a map, we cannot tell whether a tool expands us or empties us. "Skills" is too vague for a transition as large as AI.
Model
Research in progress: draft a living map—perception, world-modeling, craft/physical skill, caregiving, moral judgment, coordination, meaning-making. For each, ask: amplifiable? automatable? necessary for legitimacy? necessary for resilience?
This is a research program, not a complete atlas.
Limitations
Maps freeze moving targets. They risk Western professional bias. Figuora publishes this as an incomplete working artifact meant to be corrected by evidence, not worshipped as a framework finished product.
Assumptions
- Capabilities are plural; IQ-like scalars are insufficient.
- Some capacities are complements to tools; others are substitutes.
- Civilizational resilience requires fallback competence, not only peak automation.
Evidence
- Automation literature documents heterogeneous complementarity across tasks.
- Skill fade observed in over-automated control environments.
Predictions
- Generative AI will pressure modeling and drafting layers first; care and legitimacy judgment slower.
- Maps that ignore embodiment will misforecast robotics impacts.
Limitations
- Taxonomies multiply endlessly without discipline.
- Cross-cultural capability valuations differ.
- Measurement instruments remain immature.
Counterarguments
- Capabilities are situational performances, not inventory items.
- Focus on human uniqueness is species vanity.
Sources
- 1. Range — David Epstein (2019)
- 2. Thinking, Fast and Slow — Daniel Kahneman (2011) · Cognitive dual-process research; contested replications in places.