The agent was never alone: SIMEA

A research project I keep in the open. Classical economics models the agent as a rational maximiser; behavioural economics keeps the maximiser and catalogues its failures. SIMEA starts elsewhere: the agent as a decision subject, embedded in and steered by systems. Algorithms, platforms, recommender architectures. The framework proposes three measurable constructs: the influence gradient a system exerts on a choice, the autonomy threshold that remains the agent's own, and the entropy of the choice environment itself.

A second note takes the framework seriously and asks what it implies for forecasting: if determinism holds, does it follow that financial futures are predictable? It argues that it does not, and that predictability is a function of autonomy.

A framework written down is easy to agree with and hard to test. So the programme now has five agent based models beside it, each one running in the browser. A Schelling board where a mild preference settles into a severely separated city. A recommender that decides what a population sees, run beside the same population with a neutral feed, so that the influence gradient, the autonomy threshold and the entropy of the choice environment can be defined on screen and computed for the model while it runs. A contagion held fixed while the shape of the network under it changes. A ring road that jams with nothing blocking it. And a pile that holds itself at the edge, where the size of the next event is genuinely unforecastable. Each one cites its source and ends by saying what it does not show.