Programming AI Guided Language

The bridge between narrative human thought and formal machine structure—the language of thinking where humans teach machines how to reason, not just what to execute.

The Language of Thinking

Humans think in rich, flowing narratives; machines require precise, structured logic. PAGE (Programming AI Guided Language) serves as the definitive bridge, translating the fluid, contextual way we reason into a rigid, executable framework that AI can follow without ambiguity. Using Human-Cognitive English (HCE), PAGE encodes cognitive patterns, reasoning rules, constraints, and decision-making policies in plain, readable English—enabling persistent, inspectable, and modifiable AI reasoning across sessions. This is not traditional programming: where code executes tasks, PAGE defines how to think.

Three Core Pillars

1. Human-Readable Syntax for AI

Fully English-readable, COBOL-inspired syntax that makes reasoning transparent, editable, and accessible to non-engineers—no symbolic complexity, just clear, narrative-driven logic.

2. AI-Native Primitives & Pipelines

First-class cognitive constructs like BELIEF, GOAL, HABIT, UNCERTAINTY, and CONTRADICTION embedded as composable, reusable reasoning blocks that mirror human thought processes.

3. Simple Formal Method Assertions in English

Rigorous, verifiable logic expressed in plain language—enabling mathematical strictness and auditability without sacrificing human comprehension or narrative flow.

Syntax Reference: COBOL-Style Clarity

Borrowing from the reliability and readability of COBOL, PAGE uses structured English with reserved keywords to define reasoning patterns. Example: the HABIT primitive for default reasoning policies:

HABIT: RespondToNavigationRequest
  WHEN: UserIntent IS "Navigation Request"
  GOAL: ProvideClearDirection
  FIRST: ConfirmLocationContext
  THEN: EvaluateRouteOptions
  TYPICALLY: SuggestShortestPath
  ALWAYS: VerifySafetyConstraints
  UNLESS: UserSpecifiesPreference
  CONFIDENCE: 0.95
  BECAUSE: Users prioritize efficiency unless stated otherwise
                

All keywords are reserved; fields are structured yet narrative-friendly, enabling both human editing and machine interpretation.

Cognitive Primitives (First-Class Constructs)

PrimitiveDescription
BELIEFA proposition held with confidence (0.0–1.0)
GOALA desired outcome with priority and success criteria
ASSUMPTIONA temporary belief subject to revision
HABITA reusable reasoning policy with triggers and exceptions
CONSTRAINTA hard boundary that overrides other actions
PREFERENCEA soft bias influencing default behavior
UNCERTAINTYExplicit unknown with graceful degradation actions
CONTRADICTIONConflicting beliefs triggering clarification or revision
REVISIONBelief update with causal justification
JUSTIFICATIONStored explanation for reasoning decisions

Core Principles

Project Status: Design Phase

Current State: Evolving Specification / Active Design

PAGE represents a fundamental shift: the system design phase is the real intelligence. Coding is merely the implementation layer. We are defining how machines should think before building how they act. Syntax, primitives, and cognitive models are actively refined to optimize human-computer interaction. This is not a finished product—it is a proposal for a new paradigm in AI reasoning, where human narrative thinking meets machine-structured logic through transparent, auditable, and modifiable design.

In-Process • Not for Production Use • Specification Subject to Refinement