OS X

My office is not a starship, or why speech recognition is not for me

For most of my life, I dreamed of interacting with my computer the way the crew of the Enterprise did on Star Trek. No keyboard or mouse—just tell the computer what information you need or what task you want it to accomplish, and the right thing happens. Twenty years ago, when Apple released the Macintosh Quadra 840AV (with an extra signal processor to facilitate speech recognition), I thought my dream was well on its way to coming true.


 

But over the last two decades, every time I’ve tried working with a speech-based interface—whether it’s Apple’s long-standing Speakable Items (now in the Accessibility pane of System Preferences), the Dictation feature introduced in Mountain Lion (see the Dictation & Speech preference pane), or a third-party tool such as Nuance’s Dragon Dictate (4 of 5 rating)—I encountered a problem I couldn’t get past: my coworkers.



I wasn’t concerned about their conversations interfering with the accuracy of speech recognition; that problem has technological solutions. The issue was that talking to my computer all day invariably distracted people nearby who were trying to concentrate on their own work. (Never was this more true than when working in the home-office space I shared with my wife.) And for my part, I disliked the ease with which others could eavesdrop on everything I dictated.

Apple’s Speakable Items

Now that I have an entire home office to myself, complete with a functional door, I imagined that I could finally start talking to my computers with impunity. But I made a bewildering discovery that foiled my plans: My home office is not a starship. It turns out that voice control is a poor fit for the type of work I do here on Earth, especially given the design of OS X. And even the isolated and technologically feasible task of dictating text (which should be a great fit for an author) is unsuited to my personality type and work style.

‘Click that gear-shaped thingy. No, the other one!’

The Apple TV was designed to be operated by a 7-button remote control. The iPad was designed to be operated by finger taps and gestures. And the Mac was designed with the assumption that every user will have a pointing device (such as a mouse or trackpad) and a keyboard. A mouse would be useless with an Apple TV or iOS device. On the other hand, although Macs can be made to work with touchscreens (after a fashion), the experience is awkward because fingers don’t have the precision of mouse pointers. Operating systems work best with the input and output systems they were designed for.

Sure, you can control some aspects of a Mac with your voice, but the problem is that this capability was an afterthought, not a fundamental design choice. Many activities can’t be controlled by voice at all, and some require effort disproportionate to the task. If you ever have to tell your Mac to “click a button” or “press a key,” for example, it’s obvious that voice control is putting a round peg in a square hole.

But wait! What about Siri? It’s fantastic on iOS devices, and vastly more powerful and flexible than Speakable Items. Wouldn’t the whole solution be for Apple to add Siri to OS X?

Well, no. Don’t get me wrong; Siri and I have some great times together, and Siri on OS X would certainly improve my productivity. But even a more-advanced Mac version of Siri would only be a shortcut for a few of the things I can already do with keyboard and mouse. Siri wouldn’t help me outline a book, design spreadsheets, draw diagrams, edit audio tracks, or do any of a hundred other tasks I use my Mac for every day. It wouldn’t change the fundamental nature of my Mac or eliminate the need for existing input devices, which I would continue to use for the bulk of my office activities.

Think before you speak (or type)

I accepted the fact that voice control was not going to be useful for me, but I still thought dictation ought to be a winner, given my occupation. Sure enough, the mechanics of it work pretty well. I can get my Mac to type what I say with astonishing accuracy, and Nuance’s $160 Dragon Dictate even supports the ability to edit text by voice. Technologically, it’s fine, and without a coworker in the same room, I can even dictate without disturbing anyone. It’s just that, much to my surprise, I can’t stand writing by dictation.

What I’ve discovered through trial and error is that, for me, putting words together for delivery by reading is a different process than putting words together for verbal delivery. When I’m preparing for a live presentation, I work through what I have to say by speaking out loud. (That’s when the office door comes in really handy.) Somehow, writing or typing just doesn’t give me the result I want.

On the other hand, when I’m writing a book or an article, composing by speaking rather than typing or writing drives me crazy. I rarely come up with fully formed sentences before I start writing, and I revise extensively as I type. So, dictation, for me, is deeply frustrating. Besides, I like typing. I can do it quickly, accurately, and quietly, and I can use a large array of keyboard shortcuts, macros, text-expansion snippets, and other such tools to make the process highly efficient. It’s a better match for the way I think.

Listen to the voice of reason

Nuance’s Dragon Dictate

Many people who use voice control or dictation do so out of necessity—for example, due to injuries that make typing painful or impossible. And I’m relieved to know that if I should ever fall into that category, technology would make it possible, if not pleasant, to get my work done. I also know people who regularly use dictation by choice because they genuinely enjoy and prefer it. More power to them!

I wanted a voice-controlled computer because I thought it would be cool, but in my case, the facts didn’t match that vision. The extent to which voice control or dictation is the right choice for you depends on how your brain is wired, your office environment, and tasks you have to accomplish. As for me, I’m glad the option is there, but I’m also glad I don’t have to depend on it. Besides, I’m still calibrating my office’s warp drive, and I wouldn’t want my neighbors overhearing me. They might get the wrong idea.

Subscribe to the MacWeek Newsletter

Comments