There are lots of projects like this today; but I think the most important and useful thing (profoundly missing) is an open source state-of-art offline (or mayb

There are lots of projects like this today; but I think the most important and useful thing (profoundly missing) is an open source state-of-art offline (or maybe hybrid) speech recognition package. Until that happens, almost nothing has changed.