I’m beta-testing Windows 7, and man, the speech-recognition training program is creepy. Instead of giving you something cool to read and letting you figure out your own mistakes and corrections like Dragon NS, it puts you through your paces by reading canned lines.
- If you don’t read the canned lines, the recognition app doesn’t understand you. It’ll just say “what was that?” until you read what you’re told.
- Even if you read it perfectly, if the app is teaching you to correct a mistake, you WILL make a mistake. Which you then have to correct in exactly the way they prescribe, even if there are multiple ways to do the same thing.
- As soon as you figure out that the app isn’t really listening to anything you’re saying, a little pop-up window tells you that even though the screen doesn’t register your words, THE APPLICATION DOES. In fact, it’s using your speech patterns to program the recognition engine. So if you say “this sucks,” instead of “this is awesome,” it’ll somehow try to figure out what accent you have where “awesome” sounds like “sucks.”
- This leads to the creepiest part. The text you read is all brainwashy. It’s all “speech recognition works great! I can speak faster than I can type. I really want to do this more often.” I kid you not.
I should add that the speech recognition itself does work pretty well and the visual integration with the OS pays off. MS clearly thought hard about accessibility. However, they didn’t think at all about personality or humor.