Phil Fox Rose

« The effects of VDT radiation | Main | Designing computers as if the planet really mattered »

September 10, 1990

Voice technology

Voice technology is like networks and UNIX; it has been an emerging technology fort least a decade now.

A lot depends, though, on what people mean when they talk about computers and voice. Many think of the computer in "Star Trek", which carried on a conversation with anyone who walked up to it. Others think of an executive sitting at her desk, dictating a letter to a computer-secretary.

Since neither the technology nor the computing power is here yet to make these visions reality, voice technology developers have concentrated on systems that learn a very limited list of words spoken by a specific person. In the past, these systems have been of arguable benefit, since typing the command was usually quicker and more accurate.

A new category of voice products is emerging, though, which promises to make voice usable today and a major force in the future. These new products are based on graphical user interfaces, and integrate voice as one more object or means of control into an environment that is much more flexible than earlier systems.

Voice annotation of electronic mail is a standard feature on the Next, which has a sound input and digital signal processor. Farallon Computing Inc. has offered a simple sound input device for the Macintosh for several years.

The future direction of voice products is best exemplified by the Voice Navigator II from Cambridge, Mass.-based Articulate Systems Inc., which adds a new dimension to computer control.

Rather than sitting there-one step closer to being a vegetable — talking to the computer, the user combines typing, using the mouse and speaking to the computer. For example, in PageMaker, the user might be moving a paragraph and – while the paragraph is selected – utter "first." This instructs the computer to make the style of that paragraph that of first paragraphs in stories.

This saves time and effort in the same way that hot keys do. But it also executes a command without taking the user's attention off of the work at hand to pull down a menu or enter a special key combination. To stop and remember that that style change is Command-Shift-F7 is a lot more disruptive than saying "first." After all, we already know how to say what we want. This underscores the main advantage of voice control: It is already the way that we work.

A recent study at the University of Virginia found that MacDraw users experienced an average 20 percent to 25 percent increase in speed using the Voice Navigator and mouse over mouse alone - though it did not allow the mouse-only subjects to use macro hot keys.

The need for voice control is actually greater in graphical environments than in command-based ones, since actions are often buried several levels deep in menus.