Shafqat M. is presently Head of Content at Locate Translate UK.
AI’s advent has transformed all major industries. Its effects, good and bad, are felt by us on a daily basis.
When we’re talking to Siri or Alexa, we’re actually interacting with AI. Our transaction through payment apps or when we sift through personalized news feeds, it’s basically AI at work.
Similarly, when we use Google to translate words and phrases from one to another language, the service is made possible by artificial intelligence.
Simply put, industries across domains make use of AI. So does AI in the translation industry.
It has catapulted AI translation into a force to be reckoned with. Moreover, translation apps have become a common resort to translate words, phrases, and signs.
Deep Learning Machine Translation: “Google Translation”
The threshold is not in the viewing when it comes to AI’s capabilities.
Has AI reached its pinnacle as a translation powerhouse is a matter of debate? But certainly, it has become more accurate compared to years ago.
The impressive utilization of AI in tools like Google Translate offers insight into the unbound potential of AI in the translation industry.
Though still not perfect, Google Translate is a sneak into AI’s untapped potential and shows what it could mean if its true potential is harnessed.
Google deploys neural machine translation (NMT) in Google Translate.
Considering the rising demand for translation, what AI does is revolutionary. In business translations where it can produce translations in a matter of minutes have empowered businesses and disrupted the translation industry.
Is it quite there where it can replace manual translation completely, can’t be said. Though it surely has made progress in that direction by leaps!
Today, it has become usual for businesses to submit a machine-translated first draft of the text and improve upon it by human translators.
Language service providers also have resorted to it for the first draft, which significantly has sped up the process. Cutting at the same time the number of hours translators had to put into typing the manuscript.
Is AI Useful for Translation?
Can deep machine translation render manual translation irrelevant?
The answer is “no”.
While its usefulness isn’t doubted, it’s not quite there where it can replace human translators. As of now, the deep learning machine translation has an accuracy rate of 60-90%.
Though extremely sophisticated, the margin of error is discouraging and bars many from using it for official purposes. This much room for errors renders it unfit for, say, medical translation where patients’ life is at stake.
A single error can easily snowball into a major crisis, and machine translation is prone to errors.
Imagine a scenario where you’ve translated instructions manual using AI tools like Google Translate, and have a word or phrase mistranslated.
These documents will pay the way for wrong instructions and impede significantly the functioning of your product.
Translation agencies like Locate Translate have been using AI for an array of purposes where it has been at the forefront to automate recurrent processes such as providing translation quotes, billing, and bringing automation to the customer service and workflow processes.
AI and Post-Editing Machine Translation
The first draft we discussed above, it’s generated using AI. But the translation process as such doesn’t stop there, we’ve to proceed a step further.
The fact that AI translation is riddled with errors - remember 60-90% accuracy rate - and needs editing before it’s produced before officials or consumers, AI has given rise to post-editing machine translation.
It’s safe to say robots aren’t snatching translation jobs from humans! Contrary to what was anticipated, deep translation has given rise to post-editing machine translation, which by the way is a prerogative of humans, not machines.
Indeed, machine learning has made software translation a lot more accurate but not accurate enough to obliterate human intervention.
Whereas superfast AI translation has come in handy when traveling abroad, complete reliance on it in business and official settings is still in infancy as it’s far from being error-free or accurate enough to serve as a substitute for certified human translations.
Is machine translation useless?
No, of course, not. A step further your machine-translated text will be as good as the human-translated text.
The process goes like this. Send over the machine-generated text to a translation agency offering post-editing machine translation to make it error-free.
It’ll improve upon machine translation draft qualitatively to render it fit for official use or business documentation purposes.
AI’s capabilities fall short when factoring in localization needs, or contextualizing the translation keeping local cultures in view.
Translated texts have to incorporate local elements so that it’s relevant and complies by local cultures and regulations.
If there’s one aspect where AI beats human translation it’s price.
While a human translator may charge £0.10 to £0.16 per word, machine translation is absolutely free of cost. Therefore, prompting many businesses to use it.
Big Data and AI in Translation
Far from rendering the translation industry obsolete, Big Data and AI have instead proved a blessing in disguise for the translation industry.
With vast amounts of data to glean valuable insights from at its disposal, the translation industry has emerged as a powerhouse of analytics reports and demographic surveys, which is highly helpful in foreign market research.
The innovations AI has brought into the translation industry have added value and made it technically sophisticated, which certainty has helped translators gain, not lose value.
Key Insights on Deep Translation
Deep learning continues to transform machine translation and strives to bring 100% accuracy to it.
That’s one of the reasons why the global machine translation market has risen to $450 million in size.
However, the rising demands for localization services have put its capitalization in check, and given impetus instead to the translation industry, which was worth $46.9 billion in 2019, expected to reach $56 billion by 2021.
Besides the market, there are other areas where deep translation has been a key factor in revolutionizing the role of AI in the translation industry. It has to do with a single system mastering both source and target language.
As a result, these translation tools running on neural network architectures can instantly translate the bulk of texts with approximately 60-90% accuracy.
From what we’ve seen so far, the top-quality, contextualized translation continues to be the prerogative of humans and AI is far from overtaking it, at least as of now.
Its steady rise entails for human translators to adapt to the changing technological landscape and master it the human way.
You don’t have to lose what you’re good at just because some technology can do the same thing much quicker and at no cost at all.
Precedence lies with you, not with machines. That’s why it’s you who’s being called on to improve texts generated by AI-assisted translation tools in what is referred to as Post-Editing Machine Translation.
Furthermore, context still rules the roost, which is why localization continues to determine translation industry market size.
It’s something machines can’t master so easily. You’ve got the upper hand here compared to machine translation.
This article is accurate and true to the best of the author’s knowledge. Content is for informational or entertainment purposes only and does not substitute for personal counsel or professional advice in business, financial, legal, or technical matters.
© 2021 Shafqat M