The Collective Intelligence Observatory is the flagship instrument concept within the Algoplexity research programme.
Its goal is to make collective intelligence measurable as a real, observable quantity — by capturing the structure of interaction and the dynamics that produce it.
Collective intelligence is often described in terms of outcomes: a group solves a hard problem, a crowd estimates accurately, a team adapts quickly. But these descriptions treat intelligence as a black box.
CIO takes a different approach. It asks:
Can we measure collective intelligence as it forms, rather than inferring it after the fact from outcomes?
The answer proposed here is yes — if we look at the right signals.
The CIO framework is grounded in the Cybernetic Intelligence (CI) synthesis, which treats collective intelligence as having two separable, measurable components:
Who interacts with whom, and how? This is captured as a dynamic interaction graph — a network representation of the group where edges represent actual coordination events, weighted by frequency and recency.
When a group is genuinely coordinating, its interaction graph should become more organised over time: denser in the right places, sparser elsewhere, and more compressible as a whole.
How does that structure form and stabilise? This is captured through a motion field — a spatial representation of proximity, movement, and co-location patterns across the group.
The motion field acts as a continuous signal that feeds the interaction graph: when people move together, stay close, or co-orient, the graph edges strengthen.
A CIO instrument estimates collective intelligence in real time by:
This is a continuous, live estimate — not a post-hoc analysis.
The framework distinguishes four coordination states:
| Regime | What it means |
|---|---|
| True coordination | Interaction structure is stable, organised, and compressible — genuine collective intelligence is present |
| False coordination | Apparent structure exists but does not persist, generalise, or predict outcomes |
| Fragile coordination | Structure is present but highly sensitive to perturbation — easily disrupted |
| No coordination | Interaction topology shows no compressible pattern — the group is not collectively intelligent in this context |
These regimes are not just labels. They carry different predictions about group performance, resilience, and adaptability.
Compressibility is the key measurement concept.
A random interaction graph — one where people interact without structure — has high complexity and low compressibility. A genuinely coordinated group develops recurring, predictable interaction patterns: subgroups, leaders, channels. That structure can be compressed. It contains less surprise.
This gives us a measurable, falsifiable proxy for collective intelligence:
If a group is genuinely coordinating, its interaction structure should become more compressible over time.
If coordinated groups do not show increased compressibility, the theory is wrong.
CIO is an instrument concept in active development. The theoretical framework is established; the architecture is being refined; prototyping is ongoing.
This public repository will be updated as the work progresses and new artefacts are reviewed for publication.