Hidden Information Labs Institute studies how the structure of a network shapes cognition, behaviour, and collective outcomes. What spreads and what stalls. How closure hardens into an echo chamber. How fragmented information in health and economic systems produces failures that nobody inside the system chose.
A position in a network places demands on whoever occupies it and forecloses options they never declined. The institute calls the first structural load and the second structural unfreedom. Both are properties of the arrangement, both can be measured, and neither is visible to the person carrying them, which is why the cost usually becomes legible at the moment it can no longer be priced.
The same person performs differently in different network positions. Outcomes are treated as a property of structure first. The question is never who someone is, but where they stand.
A tie has direction, weight, and a channel that carries information, work, or access. These quantities can be mapped, audited, and changed. Constraint is a number, not a feeling.
Systems break when a channel degrades and can no longer carry the right information to the right place. Repair happens at the edges. A second path turns a load-bearing node into an ordinary one.
From a single ego network to a national health system, the object of study is the same object: a graph, its topology, and what that topology does to the people standing inside it. Each programme states the mechanism it is testing.
Adoption depends on the shape of the network carrying it. Behaviours that need reinforcement from several sources move through clustered structure and die in sparse structure. Falsehood moves through cascades faster and further than correction does. The programme tracks which topologies carry which kinds of contagion, and which trends are artefacts of the graph they travelled through.
An echo chamber is a measurable property of an ego network: high constraint, redundant ties, few independent paths to information the position cannot already reach. Exposure to opposing views inside such a structure has been shown to harden the position it was meant to open. The programme treats belief closure as a topological state with a number attached.
A patient's care is delivered by a network of clinicians who share the patient and, frequently, nothing else. Patient-sharing networks reconstructed from administrative records vary widely in structure across regions, and the gaps between providers are where information about a person stops arriving. The programme asks what fragmentation costs and who absorbs the cost.
Power is a property of the relation and not of the parties inside it. Where alternatives are absent on one side, the exchange is priced by structure before anyone opens their mouth. The programme measures asymmetric dependence in professional and economic networks, and the options it forecloses for the party with fewer paths out.
The institute's applied framework. A social echo is a signal that leaves a position, travels the network, and comes back sounding like independent confirmation. The framework reconstructs the paths a signal took, counts how many of them are genuinely separate, and returns a measure of how much of what a person hears as consensus originates with the person hearing it. Applied to ego networks, teams, and firms.
Just as neurons connect to create a mind, individual agents connect to create collective behaviour that none of them holds alone. Each agent follows local rules and sees only its neighbours, and the group still aligns, fragments, or stalls as one body. The programme models how coordination forms and breaks across a population of agents: which coupling structures produce synchrony, how coordination changes with network density, how a small informed minority steers the whole, and where a collective settles into a state it cannot leave. Computational models of collective cognition, emotion, and decision-making are treated as the same problem at a different scale.
The work is being extended from simulation into team formation and venture creation. The Collective Flow Protocol measures coordination among real people and identifies the relational structures associated with effective collective action. The institute uses those measurements to form founding teams, build ventures around them, and invest selectively in teams that demonstrate reproducible coordination. This closes an empirical loop: measure, form, build, observe, refine. Each venture becomes a live test environment for the theory, a place to check whether the structures the protocol identifies predict coordination, resilience, and value creation.
The objective is to test whether collective coordination can be measured, deliberately composed, and used as a variable in organisational and economic design.
The theoretical core of the institute: coerced adoption across network topologies. When a position leaves no alternative, the choice made inside it was never a choice, and treating it as one misreads the system as a failure of the person. The paper formalises the measure and states the conditions under which it applies.
The programmes above describe structure. The specimens below put it in motion. Each is a multi-agent simulation that runs on the GPU: particles, vertices, or agents that follow local rules, and a collective pattern that emerges from them through consensus, coupling, and phase-space dynamics. They run live in their frames and fullscreen on request. They are models of the vocabulary the institute uses, built to make a mechanism watchable, and they make no claim beyond that.
A fabric mesh with collective intelligence: position-based dynamics on a grid of agents, each carrying height, velocity, synchrony, and energy, with a consensus layer, a multi-species layer, and an environment layer read every frame. Each vertex is pulled toward its neighbours' average by a spring whose stiffness is the coupling constant, and emergent pattern formation follows from evolutionary consensus among agents that see only their neighbours. Raise the coupling and the sheet moves as one body; the individual node still moves, but it no longer decides.
Position dynamics · Multi-species evolution · Consensus mechanisms · Emergence. WebGL 2.
A particle system exploring the phase space of several strange attractors at once, Clifford, De Jong, Hopalong, and hybrid systems, and transitioning between them over time. Every trajectory ends up in a basin. The point of the model is the return: a path that keeps arriving at the same region of the space did not choose it. Basin, attractor, and repeller are the terms the Basin Profiler uses, and this is what they look like.
Strange attractors · Chaos theory · Phase space · GPU transform feedback. WebGL 2.
A living fractal surface that morphs through the complex plane over time. Height at every vertex is the number of escape-time iterations a point survives, computed each frame and drawn as a dynamic mesh. Two starting points a hair apart, one bounded and one gone, sit side by side on the boundary. That is the cascade question restated: the outcome is decided by where on the boundary a signal starts, and the boundary is where prediction fails.
Complex dynamics · Escape-time · Fractal geometry · GPU height field. WebGL 2.
Motion is reduced by your system preference, so the models are shown as still figures.
Models run fullscreen. Press Escape or the close control to return. Tap once to pause, tap again to resume. Colour can be changed while a model runs.
Papers will be listed here as they are published.
Change the structure of relation, and you change the space of possible action. The course teaches the method: read a network, measure what its structure loads and forecloses, and repair it at the edges.
If you are interested, write to Hidden Information Labs to enrol and we will walk you through the process over email. We send you the terms for the course to read first, and your place among the first thousand is fixed when payment is received.
Self-paced · Access through the archive portal
The first 1,000 enrolments include a signed copy of The Relational Architect, the institute's first book, in a strict print of 3,000 copies
Shipping included in Canada and the US
The HILI Archive holds the lecture library, the research seminars, and the tools the institute builds to do this work. Knowledge Stewards hold access to all of it for a single one-time fee, and their support keeps the archive growing. Educational access, at your own pace.
Stewardship for a single one-time fee: access to the research, frameworks, and tools held in the archive, and a hand in what enters it next.
To become a steward, write to info@hiddeninformationlabs.io and we will walk you through the process over email.
Already a steward? Enter the portal
Educational offering. Not therapy, coaching, or professional consultation. Payment plans available.
Conversations with the researchers, investors, entrepreneurs, and artists shaping culture, mapped as a network of the ties between them. Hosted by Dr. Anna Maria Matziorinis.
Each node is a guest, each edge a shared idea or relationship. Explore the graph, tap any node, and listen to the episode there.
Listen to the podcastDr. Matziorinis founded Hidden Information Labs Institute after years of research reconstructing brain networks from diffusion imaging and analysing their topology. Studying connectivity through graph theory gave her a direct view of what happens when any network degrades: the nodes remain, the edges thin, information stops arriving where it is needed, and the system loses coherence long before anyone can name what went wrong. That neuroimaging work continues separately, as a visiting scholar at the University of Bergen.
The same structural logic holds at other scales. The institute extends that work to social, professional, and institutional networks, publishing research and teaching the method to the people whose systems depend on it.
New research from the institute, and papers as they are published. Sent when there is something worth saying.
