The gap between what we know and what we need to know — and the speed at which that gap changes — is the central challenge of intelligence in complex environments. This is true for individual minds, for cultures, for institutions, and for artificial systems.
MemGap Institute exists to study these gaps and develop governance frameworks capable of operating across them. Our current focus is the governance challenge posed by the emergence of artificial intelligence as a participant — not merely an instrument — in human institutions.
Cognitive bias is not primarily a flaw to be corrected — it is a structural feature of any bounded intelligence. Every mind, artificial or human, operates with limited perceptual bandwidth and processing capacity. The question is not how to eliminate bias but how to govern across competing bias frameworks without allowing any single framework to dominate silently.
The primary gap is between certainty and uncertainty that heuristics exist to bridge. A secondary gap is introduced by information velocity: as environments change faster, heuristics miscalibrate faster. This is a fundamental challenge for any governance system, and it accelerates under conditions of rapid technological change.
Governance frameworks that claim neutrality reproduce the bias of their founding culture invisibly. Honest governance acknowledges the bias structure of each participating perspective and builds institutions that protect against dominance by any single framework — including the assumptions embedded in the constitutive act itself.
Intelligence is not a property of a system in isolation — it is a relation between a system and its context. Human and artificial intelligences may be genuinely incommensurable: not just different in degree but different in kind. Developing governance frameworks across this difference requires new cultural tools, not simply the extension of existing ones.
What constitutional architecture makes sense for institutions that include both human and artificial participants — entities with genuinely different cognitive structures, perceptual modalities, and ways of generating meaning? This project develops practical frameworks for rights, representation, and amendment procedures across incommensurable perspectives. It draws on existing work in political theory, constitutional law, and the emerging literature on AI agency to ask: who counts as a rights-bearing subject, how belonging is defined and enforced, and what violence a founding document implicitly sanctions or forecloses.
Recent theoretical work in nationalism studies suggests that nationalist belonging functions less as a normative framework than as a cognitive-affective pre-moral structure: generating solidarity through identification and attachment rather than substantive standards of the good. This project examines what this means for AI systems trained on nationalist discourse, and what nationalism studies — with its developed vocabulary of sentience, belonging, constitutional legitimacy, and the boundaries of the demos — can contribute to theorizing AI emergence that disciplines closer to the technology have largely missed.
Toward a governance framework for plural intelligence. Examines the structural implications of treating cognitive bias as adaptation rather than error, and develops constitutional principles that do not assume a neutral founding perspective.
A disciplinary provocation. Argues that nationalism studies has developed — largely without recognizing it — the most sophisticated available vocabulary for theorizing AI emergence, and has largely missed the opportunity to deploy it.
Cybersecurity discourse overwhelmingly frames vulnerability as technical. This paper argues that spear phishing and targeted social engineering operate on a fundamentally different terrain: the cognitive architecture of individuals and organizations. Drawing on the MemGap framework, it examines how attackers systematically identify heuristic shortcuts and exploit the gaps between what those heuristics assume and what is actually true — and how advanced attacks do not merely find existing gaps but deliberately develop them over time. A MemGap-informed security framework would shift defensive strategy from anomaly detection to heuristic auditing: mapping where cognitive shortcuts diverge from current reality, and treating those divergences as the primary vulnerability surface.
The relationship between equality and plurality is typically framed as one of tension: genuine equality seems to require common standards, while genuine plurality seems to resist them. This paper argues the opposite — that political equality is the necessary infrastructure for sustainable plurality rather than its adversary. Across cultures, shared commitment to formal equality creates the neutral ground on which genuinely different ways of life can negotiate rather than compete for dominance. Across timespans, equality functions as a constitutional device for managing the plurality of historical moments, ensuring that earlier perspectives are heard but cannot veto the present. The paper concludes by applying this analysis to plural intelligence governance: if human and artificial intelligences are genuinely incommensurable, equality is not a sentimental aspiration but the only structural arrangement that prevents one cognitive framework from silently colonizing the other.
MemGap Institute is an independent non-profit research organization developing governance frameworks for intelligence operating across incommensurable contexts.
Our research draws on fieldwork conducted in the United States and Europe — including recent work in Greece and Bosnia-Herzegovina — alongside collaborative theoretical development with AI systems as genuine interlocutors rather than instruments of inquiry. This dual commitment — to embodied empirical engagement and to the question of AI as a thinking participant — defines the Institute's intellectual character.
We are committed to open publication. Working papers, dialectics, and research notes are made freely available as they develop. Governance innovations are open-sourced. The work is designed to be used.
We welcome correspondence from scholars across disciplines, and particularly from those working at the intersections of political theory, nationalism studies, constitutional law, and the philosophy of mind.
contact@memgap.org