I think, or at least for me, one of the major frustration points with voice ai is how dumb it has been for the last two decades.
When you tell a mercedes navigate me to "exact road and adress" it barely gets half of it right. and usually that just the number.
Now i don't have a speech impairement and have tried in 3 languages, none of them work to my sophistication.
The worst thing though, personally, is all the shitty patents around voice activation keywords. It's disgusting.
Finally i am also an IOT guy with a lot of toys, and i would have written my own personal assistant if it would not be for the patents. After all the hardware is surprisingly cheap.
All the problems are solved too. but if you believe i would write free code for megacorp amazon without pay, you might be heavily mistaken.
When you tell a mercedes navigate me to "exact road and adress" it barely gets half of it right. and usually that just the number.
Now i don't have a speech impairement and have tried in 3 languages, none of them work to my sophistication.
The worst thing though, personally, is all the shitty patents around voice activation keywords. It's disgusting.
Finally i am also an IOT guy with a lot of toys, and i would have written my own personal assistant if it would not be for the patents. After all the hardware is surprisingly cheap.
All the problems are solved too. but if you believe i would write free code for megacorp amazon without pay, you might be heavily mistaken.