adiwan (adiwan)'s status on Tuesday, 24-Feb-15 19:35:56 UTC
-
@ceruleanspark I knew beforehand that the speech samples are transmitted over the Internet to the speech recognition servers because they are more capable of recognizing the command correctly. In this case it creates a bad feeling. I understand that they want to make the recognition more reliable with a large and varied and real-life sample and train their recognition software with that and adjust it manually. But still it doesn't make it right.
