Its what I call the "semantic divide." This is the mountain range that semantic web robots are not (yet) adept at crossing. On one side is a plethora of electronic information made available in well-defined machine-readable forms. Over on the other side (where some of our moms hang out) there are troves of useful data. All of it has even been made publicly available, though it was never intended for anything more than a mere human to interpret. For a little while longer, those of us building aggregators can keep busy with the various formats on the machine-readable side, but soon the divide will be crossed by grass-roots tech efforts. For those that can't be asked to create machine-readable versions of their electronic data, someone else will happily do it. Yes feedyes, the grass-roots semantic web is coming.