Thursday the 27th, one week from today, will mark the start of the 52nd annual conference of the American Translators Association. I’ll be giving a talk on Saturday, in the last time-slot of the event, entitled “Training a Dragon: Using Speech-to-Text to Enhance Productivity.” It’s my first time presenting at a conference, and since the software I’ll be demonstrating live can give unpredictable results now and again, I’m rehearsing each week, to iron out all the potential nasty wrinkles.
Using Dragon Naturally Speaking to augment my typing with vocal dictation has been a tremendous boon for me, which is surprising given that as recently as five years ago there was little market for speech-to-text-input among those who were physically able to use a keyboard. But suddenly, personal “listening machines” are everywhere (this month, the novelty of the Siri voice assistant has made an otherwise minor hardware upgrade to the iPhone 4 into a big seller). They’re not just for big companies anymore, like those customer service call-answering systems that can take voice commands. As a freelance translator, Dragon has allowed me to increase the pace at which I work by 20% to 40% on average, leading to hundreds of hours saved over the years.
It’s not always easy; even after years of voice-training have made the software especially clear-eared as long as it’s my voice speaking into the microphone, it still feels like a normally obedient dog that needs a rolled-up newspaper from time to time. But the occasional frustration is worth the gains in productivity and the reduced typing strain. Not everyone who’s tried Dragon has gotten to that point with it. The goal of my presentation will be to help people decide if it’s worth the time and money for them (not cheap, at $199 plus the cost of a good headset), and I’m confident there will be several people in the audience who ultimately find that it does.
Below is a video that may or may not make it into the final version of next Saturday’s seminar. This is an actual patent I translated before I owned Dragon. When I went back some years later and dictated a new translation using Dragon, as seen in the video, it took only 102 seconds to enter 124 words, including the time needed to correct two mistakes. That’s 73 words per minute entered accurately, and I have difficulty typing over 50 words without errors in that same time.
You can follow the Conference on Twitter with the hashtag #ata52. I’ll have lots of updates for those who can’t make it (@andrewlevine). If you can, don’t be afraid to say hi. I’ve had all my shots at the vet.