Voice recognition software has been like that for ages, that is an area that hasn´t improved much. There are no huge differences between them, they will just fa

Voice recognition software has been like that for ages, that is an area that hasn´t improved much. There are no huge differences between them, they will just fail at different tasks. I remember using IBM´s Via Voice back in 2000´s to dictate my school work, and it worked flawlessly - when it worked. Small deviations from the pattern were enough to make the software fail miserably. Voice recognition is not about making it understand everything perfectly as much as it´s about providing sane failure modes.