The Future of Voice

The Future of Voice
I remember the excitement I felt when I read one of the first articles discussing in glowing terms the new pen-based computers. They promised to be "Revolutionary", "The Most Natural Way To Use A Computer", and would "Add Legions Of Computer Users". And it seemed true - the truth is that pen computing was natural. What could be a more natural and easy way of communicating than picking up a pen?

Now, seven years later, the companies that pioneered pen computing are all gone: swallowed by giants or bankrupt. Pen computing was a bust, and except for a few highly specialized applications and the occasional Apple Newton bigot, there really isn't a single general use computer being sold today that works with a pen.

Why did this happen? After all, a pen is much easier to learn to use well than a keyboard or even a mouse.

The answer is that the pen was ahead of the machines that were asked to run it. By in large, the processing power to accurately and quickly understand handwriting was not available in the size of machine that a good pen computer had to be. I remember seeing an early IBM pen computer that weighed over 10 pounds, had a passive black and white LCD screen that was nearly unreadable in bright daylight, had a battery life of just under 1 hour, and had a 386SX processor inside. Talk about a dog! That machine is probably a doorstop now, which is about all it was good for when it was first assembled. I am certain that even today's fastest notebook would be challenged to read my full-speed handwriting. Heck, I am sometimes challenged to read my handwriting.

Is this what future of voice recognition looks like? The voice functions of Merlin are truly impressive - and are way ahead of where they were just a few years ago. To have a product work with generic sound cards and be able to fairly well navigate a computer without training was impossible just last year. But yet - in one month - a commercial operating system will include this feature as a major selling point.

A coworker of mine shared this story recently with me. He works with a gentleman who is one of the executive vice-presidents of one of America's largest and most prestigious public universities. This VP has no use for computers - he refuses to get an e-mail account nor place a computer on his desk, because he says that he has no time to learn it.

However, this same person did his undergraduate studies in audiology, or the study of sound and speech. Every so often my coworker gives this VP a demo of where voice recognition technology is today. The VP is ready - ready any day that he believes the technology is where he could put it to effective use. Still, he waits.

Imagine the uses for a corporate president or VP like him. Instead of having to learn to use a computer, the president could pick up a microphone and say, "Show me last month's sales figures", or "Compare this months production cost to last and graph the results". What a revolution that would be!

We are not there today, and my biggest fear is that we will never get there if people are scared off from early, unperfected voice technology like they were with pen computing. Kudos to IBM for establishing and sticking with a minimum processor and speed to use voice dictation. Only over time will these technologies become perfected, and I welcome Microsoft and other vendors to compete to find the best solution, because I believe that the spoken word could prove to be even more powerful than the pen.

John McDonald