The article discusses how rational decision-making involves using relevant information efficiently, but in reality, information is incomplete and our brains are not perfect processors. Humans tend to avoid risks rather than maximize utility. Decision-making in the real world is complex and often not achievable in a completely rational way. This leads to the concept of bounded rationality. The amount of information used for decision-making and the complexity of the decision model are related. The article suggests that machine rationality is subjective, meaning it varies depending on the situation.