# John Koza **Domain:** Evolutionary Computation / Automated Invention **Doc Type:** Thinker Node **Maturity:** Developed ## Definition **John R. Koza** is a computer scientist associated with Stanford who developed and popularized genetic programming as an approach to automated program generation and invention. ## Nexus Context Koza framed genetic programming as Darwinian search over executable structures and documented results comparable to patented human designs. [[articles/The Evolutionary Roots of Silicon Valley|The Evolutionary Roots of Silicon Valley]] connects this intellectual lineage to NASA Ames's evolved antenna work. “Human-competitive” is a performance category, not a statement about consciousness or legal authorship. A result may match an invention while remaining dependent on human-defined primitives, simulations and criteria. ## Continuity Context Koza's work makes ancestry inspectable in program populations. It also demonstrates that lineage can preserve function while radically changing form—an important analogy for substrate transition, but not proof of personal continuity. ## Sources / Provenance - Stanford EE380, “Routine Human-Competitive Machine Intelligence”: https://web.stanford.edu/class/ee380/Abstracts/041124.html - Koza, _Genetic Programming_, 1992. ## See Also [[wiki/Genetic Programming|Genetic Programming]], [[wiki/Evolutionary Algorithms|Evolutionary Algorithms]], [[wiki/Evolved Antenna|Evolved Antenna]], [[wiki/Automated Invention|Automated Invention]]