Desire-Capability Landscape

Overview

A framework for mapping AI agent opportunities by plotting worker automation desire against technological capability. The resulting quadrants identify where to deploy, where to invest, and where to proceed with caution.

The Four Zones

ZoneDesireCapabilityImplication
Green LightHighHighPrime deployment candidates. Workers want it, technology can deliver it.
Red LightLowHighCaution zone. Technically feasible but workers resist, may indicate social or quality concerns.
R&D OpportunityHighLowInvestment targets. Workers want help here but technology isn’t ready.
Low PriorityLowLowDeprioritize. Neither wanted nor feasible currently.

How It Works

  1. Assess worker desire through structured surveys that explore task familiarity, enjoyment, job security concerns, and desired collaboration level
  2. Assess technological capability through expert evaluation of current AI systems’ abilities
  3. Plot tasks on the two-axis landscape
  4. Identify misalignments between where investment is flowing and where workers want help

Why This Matters

The framework surfaces critical gaps. In the WORKBank study, 41% of Y Combinator AI companies mapped to Low Priority or Red Light zones, indicating investment disconnected from worker needs.

It also reveals that worker desire and capability are not strongly correlated (Spearman ρ = 0.17). What AI can do and what workers want are related questions, but the answers don’t automatically align.

When to Use

  • Prioritizing AI product development roadmaps
  • Evaluating enterprise AI adoption strategies
  • Identifying underserved automation opportunities
  • Understanding resistance to AI deployment

Limitations

  • Worker desires may be influenced by limited AI exposure
  • Capability assessments are point-in-time snapshots
  • The framework doesn’t capture task interdependencies
  • Cultural and organizational context shapes both desire and feasibility

Related: 07-atom—human-agency-scale, 07-atom—worker-automation-desire-stats, 07-atom—worker-centered-ai-development-question