While spatial intelligence is rapidly emerging as a foundational paradigm in artificial intelligence; its constituent ideas are by no means new. Concepts related to space, geometry, structure, and physical interaction have long existed across multiple specialized research communities, each developing its own assumptions, abstractions, and technical formalisms. What is new is the growing realization that these previously fragmented components must now be understood as parts of a unified and evolving whole, rather than the appearance of spatial reasoning per se. This chapter revisits the notion of spatial intelligence as a moving definition, one that cannot be fixed within a single discipline, representation, or task setting, but must instead be understood through its cross-cutting roles across modern AI systems.
Chapter 2