The document discusses probabilistic geological modeling and understanding uncertainty in reservoir models. It argues that current modeling tools only sample uncertainties within a given conceptual model, rather than exploring different conceptual models. Considering multiple conceptual models that fit the data is important for understanding high-impact uncertainties. Money and effort may be better spent exploring a wide range of conceptual models rather than exhaustively quantifying uncertainties within a single model. Understanding uncertainties requires interdisciplinary teams that generate different conceptual scenarios spanning realistic possibilities, rather than just investigating multiple equally probable realizations of a single concept.