SYSTEMS ATLAS
Evidence-backed representations of consequential systems
Systems Atlas asks what is actually happening inside an important system, then builds a representation that can be challenged, revised, and used to guide further investigation.
Research cycle
Representation is a working object, not a finished framework
The research repository is the working laboratory. This website is a curated public snapshot: it exposes the current representation, uncertainty, revision history, and provenance without publishing every raw research artifact.
Active system · 01
AI, Work & Organizations
What happens when increasingly capable AI enters organizations whose actual work is distributed across people, software, rules, evidence, authority, markets, institutions, and physical reality?
A coupled, adaptive, stateful sociotechnical-economic system whose constraints and workflow configuration may themselves be endogenous.
Enter AI, Work & Organizations →- AI changes the feasible set
- configuration changes
- constraints and outcomes emerge
- the system learns and reconfigures
How to read the research
Three status dimensions stay separate
Epistemic status
What the available evidence currently supports.
SUPPORTED · INFERRED · HYPOTHESIZED · UNKNOWNArtifact lifecycle
Whether a proposition remains active or has changed.
ACTIVE · NARROWED · REJECTED · RETIREDResearch stage
Where the empirical program currently stands.
SEALED · DISCOVERY · PRE-OPERATIONALMethod in one screen
Bound → represent → discriminate → test → revise
The goal is not conceptual completeness. The representation should become simpler when evidence allows, and important causal claims should expose themselves to prospective prediction, rival mechanisms, intervention, and falsification.
Read the research method →Revision principle
A good map is not protected from contradiction. It should change, narrow, or disappear when evidence warrants it.
See how the map changed →