State can be behaviorally produced
Incentives, authority, control, and burden can shape what information gets recorded, disclosed, or verified.
Systems Atlas / System 01
AI, Work & Organizations
A coupled, adaptive, stateful sociotechnical-economic system whose constraints and workflow configuration may themselves be endogenous.
Central question
What happens when increasingly capable AI enters organizations whose actual work is distributed across people, software, rules, evidence, authority, markets, institutions, and physical reality?
Current answer
AI changes the feasible set and relative cost of organizational action. Realized effects depend on the resulting configuration of actors, incentives, information, authority, task boundaries, workflows, capability stocks, and external constraints. That configuration produces outcomes and distributes value and burdens; those consequences alter future behavior, work allocation, capabilities, and organizational structure. A causal arrow may be supported by credible causal identification. A workflow-specific binding constraint is treated as intervention-validated only when it is specified prospectively, directly relaxed, passes a mechanism check, and produces the predicted downstream response.
- AI changes the feasible set
- configuration changes
- constraints and outcomes emerge
- the system learns and reconfigures
Current system map
Configuration is the causal center of v2.1
- External environment
- AI capability + cost
- Feasible action set
- Organizational configuration
- Candidate / binding constraints
- Execution
- Outcome
- Value + burden distribution
- Updated expectations / capabilities / structure
This preview is intentionally linear. The full map exposes nested configuration variables, constraint origins, feedback, provenance, and selected causal status.
Guardrails against bad inference
Important distinctions
MODEL CAPABILITYORGANIZATIONAL OUTCOME
Benchmark or model capability does not determine what a real organization can reliably accomplish.
TASK GAINWORKFLOW GAIN
A faster local task can be absorbed by verification, queues, handoffs, authority, downstream capacity, or changed workflow structure.
RETRIEVALCURRENT ORGANIZATIONAL STATE
Retrieved information may be stale, conflicting, incomplete, or inconsistent with authoritative or physical state.
TECHNICAL PERMISSIONORGANIZATIONAL AUTHORITY
An actor may technically be able to execute an action without holding the legitimate right to commit it.
ASSISTED PERFORMANCEDURABLE HUMAN CAPABILITY
High output with AI does not establish unaided diagnosis, verification, recovery, transfer, or teaching capability.
VALUE CREATEDVALUE CAPTURED
Productivity or customer value does not determine which actor receives the economic gain.
PROCESS COMPLETIONTRUE ORGANIZATIONAL OUTCOME
A ticket, workflow step, or tool call can complete without the intended downstream organizational state actually occurring.
CAUSALLY IDENTIFIEDINTERVENTION-VALIDATED CONSTRAINT
A causal relationship can be identified without proving that the same variable is the binding constraint in a specific workflow.
OBSERVED PROBLEMBINDING CONSTRAINT
A visible limitation is not a binding causal constraint until the stronger workflow-specific evidence standard is met.
HIGH CASE VOLUMEHIGH INDEPENDENT EXPERIMENTAL N
Many cases handled by a few persistent workers or teams can still provide little independent experimental leverage.
v2.1 causal repair
Three investigations changed the architecture
The workflow graph can change
AI can move economic task boundaries, eliminate dependencies, create new ones, or transform their function.
Human capability is not one stock
Production, diagnosis, verification, recovery, transfer, calibration, explanation, and teaching can move differently.
Current research frontier
The next bottleneck is private operational evidence
The immediate bottleneck is private operational evidence from a real host × workflow configuration, not another broad literature review or another conceptual layer.
FIRST PRIVATE HOST EVIDENCEresolve true-outcome linkageresolve independent treatment units
See FIELD-01 →Go deeper