Human decision making under risk

Expected utility theory was the first substantial attempt to examine human decision making under risk. What were its short comings and how did the work of Kahneman and Tversky address these short comings?