Intelligence is the ability to reach the same goal by different means. A system is intelligent to the degree that, when its usual route to a goal is blocked, it can find another – varying its means while holding its end fixed.
This is William James’s criterion – the mark of mind is the pursuit of fixed ends by variable means – adopted by Michael Levin as the working definition for agency across scales. It makes intelligence degreed and domain-relative: every Agent has some, but about different things and to different extents. A bacterium is intelligent about chemical gradients; a person about vastly more.
Intelligence is an agent’s engine of persistence against the second law. Below a certain complexity a fixed, “static” repertoire suffices to hold a system in its Attractor; beyond it, only flexible problem-solving can keep the system there. Selection therefore favours variable means as environments grow complex – random means first, then, past a threshold, Cognition.
Intelligence is one of the three parts of the coherent self an Agent comprises – the cognitive processes that navigate a Problem Space, alongside the Cognitive Light Cone that sets the size of the goals.
All intelligence is collective. A system’s intelligence is maximised when its sub-agents’ intelligences work together across scales; enlarging a collective’s Cognitive Light Cone – letting information flow more freely between its parts – is the same act as raising its intelligence.
References
- James (1890), The Principles of Psychology (mind as the pursuit of future ends by variable means)
- Levin (2022), “Technological Approach to Mind Everywhere (TAME)” (intelligence as reaching the same goal by different means, across scales)