Labels in camera space are difficult to use (bouncing up and down, hard to figure out what 3D position translates to a 2D position in the camera plane, etc.). We need a different method of placing 3D labels in camera space that makes them easier to use and thus more useful.