Experts Speak: Man Versus Machine in Translation
The debate on whether machine translation (MT) will make translators redundant rages on as technology becomes faster and smarter. Artificial Intelligence and deep learning technologies have made possible new models like Neural Machine Translation, which are making MT much more accurate. In this scenario, the question of whether humans will be rendered obsolete by the very machines they created looms large.
And yet, nothing that technology has to offer is an adequate replacement for human translation. You don’t have to take our word for it. We could list out 101 reasons why human translation is better than machine translation, but we felt it would be better to collate the latest goings on in the world of translation. Once you’ve had a look, you will see what we mean. For now, let’s get an unbiased overview from reliable sources.
Developments in Machine Translation
AI Innovations Increase
Google recently held a press event to show off a number of AI innovations, including a new tool to simultaneously translate and transcribe speech in real time. As CNet wrote, Google also demoed the feature, which is planned for future release through the Translate app and is currently being tested in a number of languages, like English, Spanish, German, and French. The feature will use AI for speed and accuracy, and the processing will presumably require an internet connection to access Google servers instead of being handled locally.
Read More: https://www.phonearena.com/news/Google-demos-live-transcription-and-translation-based-on-AI_id121842
Reaching low-resource language groups
Machine translation is an application of natural language processing that trains models on large collections of sentences in both the source language and corresponding translations in the target language. Such models are extremely accurate when it comes to widely spoken languages for which there is lots of data available, such as English-to-Spanish translations. But they don’t do quite so well when it comes to “low-resource languages” such as Lao, Kazakh, Haitian, Lao, Oromo and Burmese, for example.
Facebook’s new technique to overcome this, which involves combining “iterative back translation” and “self-training” with noisy channel decoding, has enabled Facebook’s researchers to create a new English-to-Burmese machine translation system despite only having limited data on the target language.
Read More: https://siliconangle.com/2019/10/16/facebook-makes-big-advances-ai-reasoning-machine-translation/
Neural Machine Translation takes it one step further
Using a new process called neural machine translation, AI language algorithms have resulted in far more precise language translations than were previously thought possible. Unlike earlier approaches to AI translation (such as statistical machine translation, which translated sentence fragments), neural machine translation translates entire sentences.
Read More: https://news.itu.int/ai-enabled-language-translation-developing-world/
For more information on how Neural Machine Translation works:
https://towardsdatascience.com/neural-machine-translation-15ecf6b0b
And Yet …
Facebook apologized … after its platform translated Xi Jinping, the name of the Chinese leader, from Burmese to a vulgar word in English. The mistranslation caught the company’s attention when Daw Aung San Suu Kyi, the de facto civilian leader of Myanmar, wrote on her official Facebook page about Mr. Xi’s two-day visit to her country …
Andy Stone, a spokesman for Facebook, apologized on Saturday. “We fixed a technical issue that caused incorrect translations from Burmese to English on Facebook,” Mr. Stone said. “This should not have happened, and we are taking steps to ensure it doesn’t happen again.”
Read More: https://www.nytimes.com/2020/01/18/world/asia/facebook-xi-jinping.html
Experts Speak: Why Human Translation Will Never Be Replaced
Now that we’ve seen the advancements in machine translation and the obvious failure of many of them, here are a few reasons given by experts on why human translation will trump machine translation every single time.
Can a tin man learn the language of the heart?
“Different cultures prescribe different words to various emotions, and words to express a particular emotion may not be found in a certain language or may have a slightly different understanding. To illustrate an example, … researchers state "Persian, for instance, uses the word-form ænduh to express both the concepts of 'grief’ and ‘regret,’ whereas the Sirkhi dialect of Dargwa uses the word-form dard to express both the concepts of ‘grief’ and ‘anxiety.’ Persian speakers may therefore understand ‘grief’ as an emotion more similar to ‘regret,’ whereas Dargwa speakers may understand ‘grief’ as more similar to ‘anxiety.’” In this example we see how seemingly same words have different and nuanced connotations, which may imply different association with different emotional states. Another example that research give is the existence of a word to describe a feeling in one language but does not exist in the other.”
- Anna Powers, first woman to be awarded the Global STEM Leadership Prize
Will Artificial Intelligence be able to match human comprehension?
“There are so many things that we unconsciously do when we read a piece of text. Reading comprehension requires multiple interrelated tasks, which haven’t been accounted for in past attempts to automate translation. The biggest issue with machine translation today is that we tend to go from the syntactic form of a sentence in the input language to the syntactic form of that sentence in the target language. That’s not what we humans do. We first decode the meaning of the sentence in the input language and then we encode that meaning into the target language.”
- Erik Cambria, an academic AI researcher and expert on natural language processing at Nanyang Technological University in Singapore.
Will the ability of Neural Machine Translation to consider the entirety of a sentence help?
“The problem is that considering the ‘entire’ sentence is still not enough. The same way the meaning of a word depends on the rest of the sentence (more in English than in Spanish), the meaning of a sentence depends on the rest of the paragraph and the rest of the text, as the meaning of a text depends on a larger context called culture, speaker intentions, etc.”
- Dr. Jorge Majfud, Associate Professor of Spanish, Latin American Literature, and International Studies at Jacksonville University
The Verdict
The verdict is in – Human Translation is here to stay. We, at Simpson Soft have seen the many comings and goings on in the business of translation, and our ISO 17100 guarantees meticulously verified translations. Every step is carefully reviewed for syntax, grammar, cultural relevance, local sentiments and much more. We know the effort that goes into this and can thus safely say that human translation remains unparalleled! We would love to know your thoughts!