A Revelation in Translation

The machine translation market is expected to hit almost $1bn by 2022

Machine translation, otherwise known as MT, is the process whereby software translates speech or text without the aid of a human. Digital translation services were once something of a joke, providing endless amusement by churning out jumbled sentences that barely resembled their intended meaning. But things have changed. Google, IBM, Microsoft and Skype are just a handful of businesses working on highly accurate MT to transcend language barriers. We have come a long way since the first rudimentary translation services, but what does this mean for translators, businesses, and the global economy?

A market that speaks for itself…

According to a report published in May, the global MT market will be worth $983m by 2022. By 2024, Tractica expects this number to rise to $2.1bn. The exponential growth of machine translation makes complete sense in a world that is getting smaller and smaller. Globalisation has negotiated the physical boundaries between nations, from superfast transportation to worldwide social networking sites. Businesses are keen to tap into the emergence of new markets, but this involves setting up a dialogue. They need to understand potential new customers, and potential new customers need to understand them. This has been a powerful driver for the development of machine translation. Couple the interest of huge companies with an influx of AI technology, and it won’t be long before we’re saying hello, hola, and bonjour to the ultimate digital linguist. Of course, perfect machine translation doesn’t exist… Yet. Currently, the market is experiencing the adoption of hybrid translation services that combine software with human quality control. Collaborative systems have been championed by companies like One Hour Translation, which has partnered with the likes of Amazon, Bing, and Google to combine Neural Machine Translation (NMT) with human expertise.

See you later, translator!

As it stands, hybrid systems are perhaps the most effective and accurate method of text to text and speech to speech translation. However, there are companies that are firmly pursuing the digital route. In 2016, Google claimed that its translation software could perform at the same level as a human when translating Chinese, French and Spanish into English. At this May’s I/O conference, Google Duplex demonstrated just how far Natural Language Processing (NLP) had come by having a completely convincing conversation with a human. NLP software that can understand the tone and nuances of speech will be instrumental in building high functioning, real time translation. Motivated by tough competition from Microsoft Translator, IBM’s Watson Language Translator, Babel and a host of other language focused platforms, these services will only improve. As a result, any company could become global, overcoming language barriers to enter international markets. B2B negotiations would be able to take place between anyone, anywhere, and drive the economy. Governing bodies could also take advantage of MT when debating international matters or when drawing up trade agreements. From a consumer perspective, accurate and accessible digital translation will improve customer experience when dealing with foreign companies. Wearable devices like ili have brought MT into every day life, democratising linguistics and fuelling a global society.

It’s good news for everyone, then, except perhaps for translators themselves. If a machine can do a job at a lower cost and higher quality, then that job will eventually become irrelevant. According to the Translators Association of China, there are around 640,000 translators in the world, made up of Language Service Providers and freelancers. While the machine translation industry is expected to reach $1b in the next four years, in 2015 the traditional translation sector was estimated to be worth $47.3bn. MT has some catching up to do.

As the world becomes better connected, the need to achieve high level translation has never been greater. At the moment, the adoption of hybrid systems means that we won’t be saying au revoir to our human linguists quite yet. However, the gradual improvement of AI in the form of Natural Language Processing could spell out the end of traditional translation. Not only will these platforms enable businesses to interact with new markets, but they will create a global society. Companies and consumers will communicate more widely and with more meaning. The business that develops machine translation that matches (or even exceeds) human ability will spearhead worldwide economic growth. The question is, which translation platform will get there first?


To learn more about machine translation, sign up for our free newsletter here.