I saw an interesting movie on YouTube today, which shows a programmer struggling to enter Perl code using Vista voice recognition. It took him 10 minutes to enter a simple five line program in notepad. While obviously not designed for the rigid needs of a programmer, it seemed to recognize the things that he did say quite readily. This inspired me, as an avid Vista user, to give the Vista voice recognition a try for myself.
I must admit, I'm quite impressed. Especially in applications set up for dictation, such as notepad or Internet Explorer, the performance even with minimal training is exemplary. In fact, the voice recognition probably spells better than I do with a keyboard. When the occasional mistakes are made, the dictation interface is extremely powerful in terms of its correction commands. While it is more difficult to talk to an application that does not expect dictation, such as Windows Live Messenger, you can still operate largely with your voice ( albeit a bit slower ). It even knows to scan things like emails and other data on your computer to build a dictionary of words that you might want to say.
In fact, my biggest problem I'm finding is that I can type my thoughts far better than I can articulate them. For some reason, it's harder for me to come up with what to say when I actually have to say it out loud.
To prove my point, the above paragraphs were dictated entirely with Vista voice recognition in notepad, copied, and then pasted into a Blogger window [including website navigation] using nothing but my voice. I was going to start typing at this point, but both voice dictation is actually staring to be kind of fun for me. Not fast... At least not yet. But still cool.