Every major programming paradigm shift has been driven by a change in what the developer specifies: assembly specifies bits, structured programming specifies steps, objectoriented programming specifies objects, and functional programming specifies transformations. We argue that large language models (LLMs) enable a qualitatively new mode: Agent-Oriented Programming (AOP), in which the developer specifies intent, constraints, and feedback—and a semantic agent autonomously determines the execution path. This position paper (i) formally defines AOP and its three distinguishing properties, (ii) introduces the Semantic Agent Framework (SAF), a seven-element formal model SAF = (G,O,C, A,D,R, F), and (iii) anchors AOP within the broader AI Civilization Transformation (AICT) research programme, mapping SAF to the unified execution formula E(Cap, Intent, Context, Control) → O.We stake out the terminology and conceptual boundaries of AOP to anchor a five-paper research series (AOP-1 through AOP-5) spanning paradigm theory, formal semantics,engineering methods, empirical validation, and governance.