I’m not willing to carry around a corporation controlled microphone on my face.
The only voice recognition I am comfortable with would run on open source software on hardware I owned and physically possessed. I have looked and this doesn’t seem to really exist.