I agree with the apps, but not the timing.
I think you're being extremely optimistic when you state that, "voice recognition...is likely to become standard equipment by the end of next year."
Telephones and Video communications over the Net are going to be a HUGE user app. For pennies you could talk to people all over the world comfortably!
This is going to take forever because it depends on the phone companies. They're not going to continue giving away bandwidth indefinitely. Once the net gets used by a lot of people for phone calls, you won't be able to get free local calls anymore. They're already screaming about this to the regulators and they have a valid point.
The existing bandwidth is insufficient for video and we won't see that for a long time in mass use, although I think that corporations with their own ATM networks will make extensive use of videoconferencing in the next couple of years. Unfortunately, most companies don't have the networking infrastructure today.
How about if you could just pick up a headset mic, and voice command while walking around your house, the entire system. For instance. Computer on. Activate Netmeeting. Call Grandma.
I think you can speak simple commands today, but it's not very useful.
How about instead, you just put this mic on and started dictating right to a word processor at like 150wpm! It's not here now, but with improvements it could be just around the corner.
Actually, it's here today, but it'll cost you some money and an extra board in your PC. However, we're going to have to wait for Merced in the year 2000 before we see substantial growth. Hopefully, the software cost will also be reduced. It certainly won't happen by the end of 1998.
The third application which would be great would be a total integration with the operating system. If voice commands could somehow intuitively know what your asking the computer to help you with. All the help function could be built into the voice system.
Computers will never have intuition. You must tell a PC exactly what you want! Microsoft could easily have the OS talk to you today, but it wouldn't be very useful because we already know how to read!
Imagine this. Load program in C drive, explain winsocket error 10045 ect. ect. If someone patented a system like that which ran all available software. Bill and company might have to sweat just a tiny little bit.
A prerequisite for this functionality would be a complete revamping of the help system of every app on the market. That's the tough part! Getting the PC to talk to you is a simple task that can already be done today. Application software developers are unlikely to improve their help systems because it's expensive and doesn't add much to their bottom line.
Check out the research department of Microsoft's web site. They're already involved in substantial voice recognition research, so we'll probably see it in the OS someday. Microsoft will never sweat as long as the guy who wrote the first PC program is running the comapny and he insists on working 14 hours each day, despite being the richest man in the world. |
