Been using this for a while. The cool thing about it is that google is using this to collect voice information for a huge database to be used for voice recognition software. By using it for a free phonebook, they are assured a huge cross-section of dialects, etc.