When learning in non-stationary environments, artificial intelligence (AI) can take one of three approaches: learning a model of the environment, learning how to make predictions in the environment, or learning how to control the environment. Each approach has benefits and drawbacks that must be considered before deciding which to use.
The first approach, learning a model of the environment, is best used when the environment is too complex for humans to model explicitly. The benefit of using AI to learn a model of the environment is that it can be used to make predictions about the environment. The drawback of this approach is that it can be difficult to learn an accurate model of a complex environment.
The second approach, learning how to make predictions in the environment, is best used when the environment is too dynamic for humans to make predictions accurately. The benefit of using AI to learn how to make predictions in the environment is that it can be used to make decisions in the environment. The drawback of this approach is that it can be difficult to learn how to make predictions accurately in a dynamic environment.
The third approach, learning how to control the environment, is best used when the environment is too difficult for humans to control directly. The benefit of using AI to learn how to control the environment is that it can be used to manage the environment. The drawback of this approach is that it can be difficult to learn how to control the environment directly.
The decision of which approach to use depends on the specific problem that needs to be solved and the resources that are available. In general, the more complex or dynamic the environment, the more difficult it will be to learn an accurate model or make predictions, so the control approach may be more appropriate. If resources are limited, the model-based approach may be more practical since it does not require as much data.
References:
https://en.wikipedia.org/wiki/Artificial_intelligence
https://en.wikipedia.org/wiki/Machine_learning
https://en.wikipedia.org/wiki/Non-stationary_environment