Human Translation vs. Machine Translation: Latest Trends

Is machine translation about to replace human translation?

Translation companies have always tried to provide the best quality translations possible. So far, using human translation seems to be the best option. However, some companies have recently announced new translation systems that would bring considerable improvements to the industry. Their promise is to provide high quality and natural sounding translations through convenient devices.

Recent breakthroughs in the real-time translation field

After adding instant image translation to its “Google Translate” service in 2012, Google launched its first headphones in November 2017. They enable users to have direct access to Google Assistant. They can listen to music; ask for traffic information; write emails and have a conversation in another language thanks to Google Translate. Google’s earphones – the Pixel Buds – are a real-time translating device that can translate 40 languages. Although its product gathered impressive press coverage, Google wasn’t the first company to develop this kind of technology.

In 2016, Waverly Labs, an American company, produced the first earpiece language translator. Since then, real-time translation has been a trend that no company wants to miss. Several companies worldwide have been developing their own devices such as LeTrans, Travis The Translator, Ili Wearable Translator and Bragi Dash Pro.

How do these new technologies compare to human translations?

These technologies are very easy to use. For most of them, all you need are earpieces and a smartphone. With the translating earphones, two people who speak different languages can both hear the words in their own language instantly during a conversation, and straight in their ears.

These technologies aim at making international communication easier. The founder of Waverly Labs reported that the idea came to him as he was experiencing problems whilst communicating with his girlfriend. She spoke French and he spoke English. Using this kind of technology was a good solution to overcome the language barrier in their relationship. It helped them have more natural conversations.

However, when Google launched its earphones in 2017, the press generally agreed that Google had failed to deliver its promise.

In fact, most of these devices are really only efficient for basic conversations. Having a more elaborate conversation via these earphones is somewhat more difficult, as some translations sound unidiomatic. However, other translating systems do exist.

Achieving human parity

More recently, Microsoft announced it had developed a translation system as accurate and idiomatic as human translation. It said it had achieved human parity on news stories translation in one language pair (Chinese to English).

Their system has been tested by bilingual human editors. They compared Microsoft’s results to professional human translations and approved them. Microsoft even wrote on its website that their system “significantly exceeds the quality of crowd-sourced non-professional translations”.

Microsoft is now working on real-time news stories translation.

Almost two years ago, Google also announced that their system’s translations was “nearly indistinguishable” from human translations.

Although machine translation has recently made breakthrough improvements, human expertise is very often still required to proofread and edit translations.

Human translation, proofreading and editing services

For the past 15 years, LingoStar has been providing translations made by native speakers and professional translators with more than 20 years of experience translating in their areas of expertise.

For more information about our services, visit our website. You can also request a free quote, call us on 604-629-8420 or email us at info@lingo-star.com.