Abstract
The Universal Optimization Framework demonstrates that systems maximize entropy production efficiency through Ω/K optimization, but does not specify how this optimization is achieved. We identify five constraints that form a complete coordinate system for describing optimization patterns: boundary definition, pattern recognition, resource allocation, integration, and temporal continuity.
These constraints, derived from the logical requirements of observation rather than assumptions about systems, are both necessary and sufficient for distinguishing organized from chaotic dissipation. Analysis of 256 cellular automata rules and 47 Game of Life patterns confirms five dimensions capture >95% variance in system behavior. Knockout experiments demonstrate each constraint's necessity—removing any single constraint reduces system viability to 6-12% of baseline. Hidden Markov Models converge to exactly five hidden states, validating sufficiency.
The constraints provide a measurement framework for observable behaviors that observers interpret as learning, memory, and decision-making. Intelligence emerges not as a property certain systems possess, but as patterns recognized when systems navigate five-dimensional constraint space while optimizing thermodynamic efficiency.