Voice recognition ups and downs

The You Tube video linked in the title is a taped live demonstration of MS Vista's voice recognition. Unfortunately, this clip was shown in many places but not in its entirety, and with a definite negative slant. I am not going to enter into the MS Office Speech-to-Text vs. Dragon Dictate debate, because I believe both have their place in the large assistive technology picture. I have used both with students/clients and had great outcomes and not so great outcomes when the match wasn't good.
I love Microsoft's tool for my students who need some support with homework and long assignments. It's easy to access, train, and most individuals already have it on their home computers. These student's often edit with their keyboard as well as their voices. On the other hand, if I have a student who has limited use of his/her hands, or needs to use speech-to-text for the bulk of his/her assignments, then I favor Dragon.

But the real intent of this post is to question why technology acts up when we present it? How many times have you been presenting and some unexpected glitch pops up? The link you've used a dozen times, or the video that ran in your power point for two weeks solid everyday leading up to the conference presentation... Why? Are there technology gremlins running amok? Do the butterflies in our stomach produce some sort of magnet field that causes the technology to crash? Or, is the technology having it's own identity crisis, since only the piece being showcased usually acts up, not the other 6 gadgets you've got running simultaneously.