General AI term
A trained computational system that produces predictions or generated outputs from input data.
An AI model maps an input to an output using patterns learned during training and whatever context the current system provides. A model does not automatically know the facts, audience or purpose behind a request.
The model matters to prompt engineering because context limits, tool access, instruction-following behaviour and output controls differ. Gaiish remains model-independent: use the structure with the model you actually intend to run, then validate the result.