Machine Translation Evolves… But How Does It Impact Translators’ Jobs?
Beijing and New York City may be 10,985 km (6,826 miles) apart, but chances are, their locals are watching the same on-demand TV shows, downloading similar e-books, and shopping from the same online retailers. But, they’re doing so in their native languages and dialects.
New Era, Marked by a Massive Demand for Translated Content
Globally, people are consuming content at an increasingly rapid pace, and as a result, the demand for translated content has reached an all-time high. From blogs to e-books to on-demand TV shows and e-commerce websites, today’s companies must translate large quantities of content — quickly — to consumers.
Fortunately, machine translation (MT) technology is more available and the quality of its output has greatly improved. This has made it possible (and more affordable) for companies to offer translated content in large volumes. Machine translation involves computers translating text into other languages with minimal or no human intervention. As a result, even the smallest companies now can afford large amounts of translated content. According to Gergely Vándor, Product Manager at memoQ,
What Does this Mean for the Translator Community?
Today machine translation usually involves post-editing work done by humans. This post-editing process ensures the content meets the highest quality level desired by the client. For translators, this means they are mastering a new skill that requires a new type of training — post-editing work. It can be challenging for translators to adapt to this new job skill, but fortunately, if they do, there are many benefits.
Machine translation can help them deliver work faster, and therefore, take on more work. It can also make their jobs easier. The technology gives translators a starting point, an opportunity to improve quality, and the ability to churn out more content quickly.
According to Arle Lommel, Senior Analyst at Common Sense Advisory, and Donald A. DePalma, Chief Strategist at Common Sense Advisory, “[Companies now] use the machine to perform the boring work of translating simple content and ensuring terminological consistency, while they route complex and interesting content to humans to take advantage of their skills.”
Prices Go Down, but Job Opportunities Go up
When machines do most of the legwork, naturally the prices of translated content go down. This is making many translators across the industry nervous. They’re asking questions like, “What’s my new role now that machine translation is readily available?” “Will I still get the same amount of work?” and “How does this impact my pay?”
Fortunately, with post-editing needs, the job opportunities are still there. But, there needs to be an efficient way to measure post-editing efforts and then pay translators accordingly. Fortunately, memoQ, among other similar translation software companies, now offer product features to help measure these efforts accurately, giving translators a unique advantage in a shifting industry.
According to Vándor, “There are two approaches: one is to measure the time it takes for a translator to translate the document to its final state. The other approach is measuring the amount of edits that were done.”
The Question is — Will Neural Machine Translation be a Game Changer for Translators?
Earlier generations of machine translation technology work like a dictionary and use algorithms to decipher grammar, syntax, and phraseology. Therefore, they still require post-editing efforts from humans. However, machine learning (also known as neural MT) “learns” how to translate languages from the information that is given to it. The more content it translates, the “smarter” the technology gets.
According to a recent article in Forbes, the role of translators has changed (and will continue to change) with a greater use of neural machine translation. (2) Neural MT reduces the amount of post-translation, post-editing work needed. But, it also gives smaller companies access to translation services, so they can request work when they didn’t have the budget to do so before. This can actually create more jobs for translators.
However, Balázs Kis, Co-Founder and Chairman of the Board at memoQ, doesn’t necessarily think neural MT is a true game changer.“ It looks like neural MT could be a breakthrough because the translations are well-formed, and for the most part, grammatical. But, this also makes it more difficult to spot missing or altered parts,” says Kis. As a result, translators are needed to ensure the output is high-quality content.
It looks like only time will tell how neural MT will impact our industry.
But this we Know: MT Technology is Here to Stay
In recent years, and most likely in 2019, the use of machine translation has grown and will continue to grow exponentially. Until recently, engineers and information technology experts spent a great deal of time setting up an MT system. It required a large investment in resources and server space. But now, cloud-based and hosted solutions make launching machine translation technology simpler, faster, and more affordable. As a result, we believe you will see even more of this technology in 2019.
There’s also a trend in integrating machine translation technology with speech recognition devices. Say hello to Alexa, Siri, and other popular products. This technology allows a face-to-face dialogue, despite language differences. For example, in 2019, you might see more machine translation technology integrated with newer speech recognition devices, as well as within older, everyday devices, like photocopiers and megaphones.
No one can deny machine translation technology is here to stay — and it’s quickly evolving in ways we never imagined. From neural machine translation to integrations with speech recognition devices, MT is evolving, and becoming ingrained in our world and how we communicate. But, there are still big questions to answer. We need to embrace the technology, but also spend significant time and resources ensuring it delivers the best quality. And ensuring there’s still a place for human translators.
References:
1 “Machine Translation: 2018” Arle Lommel and Donald A. DePalma.
2 “Will Machine Learning AI Make Human Translators An Endangered Species?” Bernard Marr.
A Desire for Actionable Analytics and Reporting
