Principlev1PR-1109
When modifying an agent, evolution is preferable…
When modifying an agent, evolution is preferable when the core architecture remains sound and the gap between current and needed state is small; replacement is preferable when core assumptions have shifted or the architecture cannot represent new requirements.
Why This Is a Principle
This is a principle derived from Complex systems that work evolved from simple systems… (simple systems that work evolve from simple systems that worked) and You can design your own cognitive processes rather… (you can design your own cognitive processes). It provides actionable guidance for the evolve-vs-replace decision based on measuring the distance between current and needed state. It's general enough to apply across many agent types but specific enough to be actionable.