Despite decades of interest and research in 3D user interfaces, most of us continue to communicate with our computing technology via primarily 2D interfaces. Even the most sophisticated software that allow artists to create dazzling 3D special effects in movies, or designers to construct sublimely beautiful 3D physical artifacts, rely on 2D interfaces to generate the 3D output. Why? Is all our research into 3D interfaces simply crummy, and hence not worthy of adoption by the broader community? Perhaps the 3D interface research community has simply failed to persuade our industrial colleagues of the value of all this wonderful 3D interface technology? Or maybe 2D interfaces are simply better for most if not all tasks, and researchers are just wasting their time exploring 3D interfaces? Or are there simply technological hurdles, that once surmounted, will finally allow 3D interfaces to flourish? I suspect that the truth lies in a combination of these hypotheses, and in this talk will probe them via examples from the past, present, and near future.