Governance begins before the decision.

Information systems do more than transmit content. Their schemas determine which distinctions survive representation; rankings and defaults structure visibility and action; interfaces define what can be perceived, refused, contested, or changed. I treat those choices as governance-relevant because they shape the environment in which later decisions become possible.

A system's ability to infer, recommend, rank, persuade, or act does not by itself establish legitimate permission to do so. A related distinction organizes the technical side of the work: verification is not validation. A record can establish what happened without establishing that the outcome was legitimate, just, or appropriately authorized.

The program draws from political philosophy, social epistemology, systems theory, cybernetics, feminist philosophy, science and technology studies, complex systems, and polycentric governance. I work from a fallibilist position: there is a reality we are trying to understand, but no observer has complete access to it and not every perspective carries equal evidence.

• Relational: subjects, institutions, technologies, and environments develop through interaction.
• Polycentric: robust systems need multiple partly independent centers of knowledge, interpretation, and authority.
• Anti-domination: concentrated, arbitrary, unaccountable power is a governance problem even when it is technically effective.
• Process-oriented: learning, revision, error correction, and developmental trajectories matter alongside outcomes.
• Situated: exposure to risk, access to knowledge, credibility, and practical standing to question are unevenly distributed.

Cybernetics is useful here as a connective tradition rather than a narrow computing specialty: feedback and feedforward, regulation, adaptation, observability, requisite variety, recursion, self-reference, viability, learning, and the role of the observer in the system being studied.

I follow those ideas across control theory, complex systems, cognitive science, organizational theory, ecology, sociology, institutional design, media and communications, design, and governance. The recurring question is whether a low-dimensional model is being given authority over a high-variety world it cannot adequately represent.

• When does information architecture become a form of private governance rather than merely communication or recommendation?
• How do attention, representation, compression, memory, and interface design distribute practical and epistemic power?
• What helps people recognize a frame as a frame, tolerate competing interpretations, and revise without collapsing into relativism?
• How do AI systems alter expectations of reciprocity, disagreement, authority, intimacy, and other minds?
• Which computational paradigms beyond current LLM scaling are important for understanding intelligence, efficiency, embodiment, and control?
• What would polycentric technical and institutional infrastructure require in order to coordinate without imposing a monopoly over interpretation?

I track world models and latent predictive learning, neuromorphic and event-driven computing, analog and in-memory systems, embodied and collective intelligence, multiscale and recursive processing, verification and validation, formal methods, provenance, and runtime governance.

Funding, hiring, infrastructure, and scientific attention are useful signals about where a field is moving. They are not evidence that the underlying claims are correct. I am interested in mechanisms, benchmarks, negative results, abandoned approaches, and what would falsify a favored explanation.

My practice combines primary-source reading, technical diligence, lectures and seminars, open technical communities, direct conversations, field research, and a multidisciplinary network built over more than a decade of work across safety-critical and sociotechnical systems.

• Separate observation from inference. Keep evidence, interpretation, hypothesis, uncertainty, and normative judgment distinct.
• Follow counter-signal. Negative results, failed replications, disagreements, field failures, and abandoned architectures matter.
• Trace structure. Ask how incentives, authority, information flow, dependency, and feedback produce the observed behavior.
• Preserve revision conditions. State what evidence should change the model.

These are working concepts, not claims that every phrase is historically unique. The contribution may lie in the definition, mechanism, synthesis, or application, and each remains subject to literature review and revision.

• Conditioned solipsismA proposed relational and epistemic orientation in which independent subjectivities increasingly fail to register as meaningful constraints while the environment producing that condition becomes difficult to perceive as governance.
• Perspectival bandwidthThe practical capacity to hold, compare, and update across multiple situated models without treating difference itself as evidence of error.
• Point of RelationThe developmental layer at which systems shape how subjects encounter otherness, reciprocity, negotiation, and shared reality.
• Governance of becomingGovernance evaluated not only by discrete outputs but by the learning conditions and developmental trajectories a system repeatedly produces.
• Polycentric epistemic infrastructureInfrastructure that supports coordination and verification while preserving multiple centers of interpretation, provenance, dissent, and revision.