Intelligence

An agent's capacity to build models whose predictions reliably match reality across a widening range of novel situations, and to correct them when they fail. Intelligence is measured, not declared -- by how broadly, how reliably, and under how much novelty and stake an agent's models keep matching reality, and by how fast it corrects error. It is a continuum, not a kind: it does not depend on substrate or on how an output was produced, only on whether the models work. The distinction is agent vs non-agent, not human vs machine.