Don't worry, I'm not going to be burning my keyboard; but that badly-expressed metaphor does lead nicely into my thought for the day - what is the point, for the average computer user, of speech recognition systems?
I'm sure there's a market for all those developers and typists who have sacrificed their wrist tendons at the 104-key altar, but this isn't a post for them.
As seductive as the image portrayed by Star Trek of talking to your computer is to the average geek (although it only seems to be Captain Picard who gets the voice recognition systems, everyone else still seems to get keyboards), there are a few warning signs:
- every time Intel brings out a new super powerful processor, pundits proclaim that this is the one that will make speech recognition a practicality (yay for alliteration!). They've been doing this since the first generation Pentiums and are still doing it, so a pinch of salt is needed;
- Optus customer support. OK, so not the most obvious reason, but if you've ever had the misfortune of having to get help from Optus, you'll know the pain of their voice recognition menu systems. I have few problems with using menu-driven destination selection when the menu is driven by your phone's keypad, but trying to tell the system which option you want is an exercise in futility. Optus refuses to understand my very English accent, and the only way I can get it to recognise me more than 20% of the time is to speak like Elmer Fudd, "vewwy vewwy swowy" and with the same accent. It's so embarrassing, I have to hide myself away even from my own family, just so I can complain about my busted bandwidth cap;
- Imagine a typical office. Now imagine 20-odd users (not 20 odd users), all trying to tell their computers what to do. Can you hear yourself think? It'll sound more like the local pub at 7pm.
So when I see yet another pronouncement that speech recognition is ready for the masses, I shake my head, and keep my mouth shut.