I think the comments listed there kind of miss the point about speech recognition and UI, assuming that speech recognition will be used to drive current GUI concepts rather than a speech-oriented UI (i.e., non-GUI). Of course I'm not going to be telling my computer "resize my window three hundred by six hundred" because... there's not going to be a window.
On the other hand, the comments about privacy and noise are spot-on. Interestingly, we have the technology to beam-form sound directly to our ears... yet we don't have that capability over our mouths. There are some experiments with sub-vocalized commands for UIs, along with the utopian "thought-control" UI (hope you can think clearly!), but speech recognition by itself will likely remain a niche UI.
|