Specifications underspecify their implementations. This paper identifies the rendering problem – the structural pattern in which a specification of bounded complexity is rendered into an implementation of vastly greater complexity, producing emergent phenomena that no specification contains. We identify three primary domains where the rendering problem manifests: biological (genome to phenotype to consciousness), organizational (org-schema to operations to customer experience), and perceptual (brand emission policy to signal field to perception cloud). In each domain, three structural properties hold: (1) a specification gap – the implementation contains substantially more information than the specification; (2) a configuration layer – the same specification produces different implementations depending on contextual parameters; and (3) irreducible emergence – the system produces phenomena that cannot be derived from any lower layer (Bedau, 1997). We ground these claims in established biology (Dawkins, 1976; Waddington, 1957; Fraga et al., 2005), process philosophy (Whitehead, 1929; Maturana Checkland, 1981; Ackoff, 1981), and non-ergodic dynamics (Peters, 2019). Six propositions are advanced, each testable within its home domain. We argue that prior organizational models – particularly Morgan's (1986) "organization as organism," Neilson et al.'s (2003) "organizational DNA," and Beer's (1972) Viable System Model – capture aspects of the same phenomenon but do not formalize the specification gap or configuration dependence as structural properties. The paper concludes that as artificial intelligence accelerates implementation speed, specification becomes the binding constraint – and therefore the locus of value – in biology, business, and beyond.
Dmitry Zharnikov (Tue,) studied this question.