I've just been experimenting with the Windows speech recognition in Windows 7. I'm able to dictate information into posts here without too much trouble.
It's pretty slick to be able to sit back and just talk to the computer and have it show up as text in the editor here.
It's a bit different to have to think before one talks instead of just typing and go back and correcting text later.
This works surprisingly well, better than I expected.
I'm also able to activate the various forum controls without too much trouble. There are a few however that are a bit sticky.
The speech recognition seems to integrate with Internet explorer pretty well… Most controls can be activated just by saying that name on the button or the link.
I seem to have figured out most of how to use Internet Explorer via voice command, but there's just one thing I can't figure out how to get it to do on the Adobe pages:
From any forum, if one clicks the Adobe Forums link (the one at the left end of Adobe Forums > Adobe Photoshop > Photoshop Windows, for example), one is taken to the Adobe Community Forums landing page. If one scrolls down, one sees a series of gray boxes, one of which says GENERAL FORUMS. I simply cannot figure out how to get the speech directed application operation feature to click on that.
Ah, speech recognition.
I am a tad worried about that. My wife's car has some major speech recognition. One can control much of the operation of that auto, via the speech input. I always worry that I might curse, while driving it, and the danged car will attempt to follow what it assumes are my commands! I can imagine all sorts of uncomfortable problems, when someone cuts me off in traffic, and I shout out a phrase, that I do NOT want the car to act upon...
Especially tell it the "go to ...." command liable to try to drive you DC or Texas.
You guys make light of these issues, but they really ARE issues in all seriousness - Imagine a misinterpreted command that causes deletion of data.
With power comes responsibility and right now computer voice recognition is both somewhat touchy and of course completely without conscience.
Basically while it's active you have to stay focused on it, and you have to remember to say "Stop Listening" when you want to do something else. I've also found that it responds to ambient noise - even keyboard clicks - though the words that are interpreted from such things haven't caused havok - yet.
Surely it's still just a toy, with possibly some utility in taking dictation, though frankly one of the reasons one dictates things is so one does not have to spend the time to ensure every word is right - and that's just not possible as it is. It gets things wrong just often enough that you have to proof-read everything carefully.
Speech recognition is not ready for prime time It need about 50 years more research to prefected before it ready. Each person has a unique way of speaking (Dialet) and some people's way of speaking. For example the telephone system in my mothers car, can not even understand my mother. She's 87 speaks low and has quiver to her voice (a combination old age and essential tremours - a mild form of Parkinson, that don't get any worse or better.) And It can't understanding say 4. I Have to draw it out the number 4 before it will understand. Or Zero I have to say a specific way.
I wouldn't trust using Speeching recognition in anything other than the Phone.