Machine Translation: The Big Tech Companies
As Google and its main big tech rivals continue to announce developments in their machine translation services, should human translators be fearful that their role will one day be completely replaced by a chatbot?
Machine Translation: A Viable Alternative
Early machine translation services were woefully inadequate, but as artificial intelligence develops in other areas, this software is now being applied to machine translation with, it would seem, some promising early results for a limited choice of languages. Intelligence is the key word here. Previously, online translation was limited by the restrictions of its pre-ordered software, whereas now it has the freedom to decipher language, mimicking much more closely the work done by a human. Machine translation, however, is still a very long way off the complex work done by translators in business or more specialist fields, such as technical, medical or political. Industry observers believe it will be many years, if ever, before a machine can wholly replace a person.
There are over seven thousand languages spoken in the world today, so it’s going to take a very long time to capture and understand all of these and the possible permutations are endless, language to language; online machine translation companies are concentrating their efforts on some of the most commonly spoken languages at the moment: English, French and Spanish.
Context, Style, Register and Tone
Despite all of the advances in machine translation, it’s never simply a question of replacing words with the correct translation in the target language. There are so many other aspects of language to consider, including, but not limited to, register, tone and readability. Context is also crucial and requires an understanding of the historical development of words and that language is never static: contextual development and relevance is constantly evolving.
Security and Safety: Deciphering the Code
Security and safety are two other very important considerations, which have not yet been fully explored or discussed, mainly because machine translation has not yet really proved itself sophisticated enough to be taken seriously by industry professionals. But imagine if a spy in the camp at Bletchley Park had decided to tinker about with the piece of code he had just deciphered. There is always a natural element of human error in terms of accuracy, which in the main part can be eliminated by reviewers and stringent quality control measures, but what about hackers? The potential for mischief or misconduct is endless and may hold very serious implications, as is likely to be the case whenever a human-to-human process is replaced by a machine.