Recent works have emphasized the role of complexity as a critical constraint on human behavior following Herbert Simon’s original proposal (complexity-cost hypothesis). By contrast, we propose, in line with recent neuroscience models, that complexity can often be appealing (complexity-value hypothesis). Complexity can attract decision makers because it is associated with a large diversity of outcomes, thus offering many opportunities for the resolution of uncertainty, which is inherently appealing to humans. Using incentivized experiments, we show that, in the absence of immediate feedback on lottery outcomes, decision makers prefer lotteries with less outcomes (low-entropy) in line with the complexity-cost hypothesis. However, when feedback is provided and opportunities for resolving uncertainty are thus offered, this effect disappears in line with the complexity-value hypothesis. We discuss various implications of these findings in human resource management, marketing, and finance.