The discipline.

Four nested layers.

The discipline is articulated across four nested layers, each concerning a different aspect of how an AI system is composed when human intelligence sits at its structural center — authority, the knowledge that authority operates within, the architecture that holds them together, and the compounding that operation produces.

1. Layer One

   Authority

   Every system architected to this discipline is organized around named human authority. A domain expert sits at the structural center. Their judgment is locatable, attributable, and defensible. We do not build systems where authority is anonymous, ambient, or unaccounted for.

2. Layer Two

   Knowledge

   The knowledge an AI system operates within is bounded by design. Its scope is articulated; its boundaries are defended; its contents are inspectable. Bounded knowledge is the precondition of accountable AI in any serious domain.

3. Layer Three

   Architecture

   The system is architected as a coherent whole — human authority, bounded knowledge, and the artificial intelligence that binds them together as a single object of design. The discipline holds that this composition, not the model alone, is the proper unit of analysis.

4. Layer Four

   Compounding

   The system grows more valuable the longer it operates. Use deepens the knowledge it works within; the named expert's decisions accumulate as precedent; the system, properly architected, compounds rather than depreciates. We hold that this property is the discipline's most distinctive promise.

Six decades of foundational work.

The discipline is built from named, peer-reviewed work — each contribution dated and attributable. What follows is the canonical list of foundational sources organized by the layer of the architecture they inform.

1. Norbert Wiener

   Massachusetts Institute of Technology

   1948

   Cybernetics — the mathematics of control and communication in animals and machines. Established the systems framing in which human and machine intelligence are studied as coupled feedback architectures rather than rival faculties.

2. J.C.R. Licklider

   Massachusetts Institute of Technology / ARPA

   1960

   Man-Computer Symbiosis — the foundational paper articulating that the productive future of computation lies in close coupling between human cognition and machine processing, with each performing what the other cannot.

3. Douglas Engelbart

   Stanford Research Institute

   1962

   Augmenting Human Intellect — the conceptual framework that defined the goal of computation as increasing the capability of a human to approach a complex problem, gain comprehension, and derive solutions. The direct philosophical predecessor to Hi-Centric-AI.

4. Michael Polanyi

   University of Manchester

   1966

   The Tacit Dimension — the philosophical account that knowledge held by named human authorities is in part irreducibly tacit, cannot be fully encoded, and must be preserved at the structural center of any system that operates within a domain. Source of the named-authority framing.

5. Herbert Simon

   Carnegie Mellon University

   1969

   The Sciences of the Artificial — established that artificial systems are designed objects subject to architectural analysis, and that bounded rationality, not optimization, governs how cognitive systems operate within domains. Source of bounded-knowledge-field framing.

Architecture, not autonomy.

Hi-Centric-AI as a discipline is the work of organizing six decades of named research into a coherent philosophy of artificial intelligence designed around human intelligence. Wiener established cybernetics. Licklider formalized man-computer symbiosis. Engelbart defined the augmentation goal. Polanyi located the tacit. Simon named the architecture of the artificial. The contribution here is integrative — naming the discipline, articulating its commitments, and extending the lineage into a working philosophy for AI in the present.

Without supplanting the lineage, without departing from its commitments, and without the autonomy-maximalist claims that define the rest of the field.