In the science-policy interface scenarios to a low-carbon future are often made using Integrated Assessment Models. But there are many more ways to explore the future, such as visioning, gaming and speculative design. Why have IAMs become such dominant tools?