This document discusses uncertainty in intelligent systems. It defines uncertainty as doubtful information that occurs due to problems with data like missing or unreliable data. It describes three issues to address with uncertain data: how to represent it, how to combine multiple pieces, and how to draw inferences. Methods for managing uncertain information include probability, Bayesian networks, and hidden Markov models. An example of uncertainty in decision making is given around driving someone to the airport. The document also discusses representing knowledge with degrees of belief and updating beliefs based on new evidence.