We live in a world where language technologies, still in their infancy, are powerful enough to turn your web browser into a text translator, and your smart phone into a speaking interpreter you can take with you wherever you go. How will our world change as these technologies reach maturity? As always when looking to the future, Ray Kurzweil is willing to share his own interpretation on how things will unfold. Watch the video below to see his short interview with Nataly Kelly on computer translation and the future of the language barrier. Kurzweil has predicted computer translators will be common by 2019, and reach human levels by 2029, but won’t many of us feel the impact of these technologies years before they reach these milestones?
Nataly Kelly, is the Chief Research Officer at the Common Sense Advisory, a firm focused on globalization, translation, and other language-influenced aspects of business. It makes sense then that her questions revolve around how computer translation will interact with, and perhaps come to replace, human interpretation. She asks Kurzweil if we will live in a society without language barriers (response at 0:40), if his predictions about translation technologies have come true sooner than expected (response at 3:08), and what lead him to be interested in the work of Franz Och well before he developed Google Translate (response at 5:05). Kelly then goes on to probe how technology will affect spoken language interpretation (response at 8:35), how Kurzweil’s own spoken language tech has developed (response at 10:50), and the ultimate value of learning a new language in a world of computer translation (response at 12:25). Her final question, will human interpreters be replaced by machines (response at 14:10) is one that both speaks to her base and is echoed in the concerns of workers everywhere. Rest easy, human interpreters. Kurzweil believes translation technology will transform human work but not replace it, as synthesizers did with music in the early 1980s.
I’ve been very impressed with the translation technology (both text and spoken) that has arisen in the last few years, and unabashedly enthusiastic about where this trend in development may be leading us. In this interview, Kurzweil comes off as taking a relatively conservative stance on computer translation, playing up the uniqueness of language, the art of translating literature, and the power of technology to augment humanity rather than replace it. I don’t argue with any of these points – certainly the subtle nuances of language may stay beyond computational comprehension for many years to come, and there will always be a value to learning a new language.
What I question is the size of that value, and the importance of language nuance. Reading Gabriel Garcia Marquez in English is different than in Spanish (honestly, I had difficulty with both) but not everything is great literature. As anyone who has shopped in a market where you have next to no language skills can tell you, you can do a lot without knowing much at all. “I want [point] that one. No, the [make hand gesture] big one. Yes. That is good robot. I buy [hold up fingers] 5 dollars.” As Kurzweil points out, we already have lower level computer assisted language translations now, and we’ll have more powerful versions soon that will be effective. That, I think, is the real message to take away. Sure, for a long time to come there will be a place for humans at the translation table, but the impact of interpretation technology will be felt long before computers reach human level skills. In fact, I’m already seeing changes in my own browsing habits and concerns. Chrome’s automated translation has expanded my web surfing to new continents, new ideas, and new cultural concerns. Imperfect, hell let’s be honest, openly flawed translations are still good enough to change the way I view the world. Forget 2019, or 2029, computer translations are already here and making their presence felt. All I can say is, to me, that is good robot.
[source: Huffington Post]