Is machine translation (MT) advancing in leaps and bounds or is it going nowhere? Who uses machine translation technology and are they finding it useful?
Machine translation technology has been around for quite a while now and one would have thought that it would have become sufficiently advanced enough that professional translation services would be taking advantage of it. That doesn’t seem to have happened and MT is somewhere in a middle-of-the-road situation. It’s advanced enough that it can be used for amateur translation and take out some of the load from the work of professional translators but isn’t yet good enough to replace human translation for anything which is technical or official in nature.
Improvements in machine translation software and technology lie with software engineers and software developers. They are in a bit of a quandary because there is a certain amount of doubt about the eventual goal of machine translation. Where is it going? What will it be able to do in say 10 or 20 years’ time?
The advantages of a fully automatic translation tool for any sort of translation services in Australia or anywhere else in the world are obvious, but can MT ever replace the depth of knowledge of a human translator?
Those who work for professional translation services do not have the time to refine machine translation vocabulary and grammar abilities. Time is money and they depend on their own skills to convert text from one language into another as accurately as possible. At present, even the best MT translations need significant post-translation editing and proofreading. Granted, that’s true for any translation task. Woe betides any translator who thinks their raw translation is good enough for a client. The better the editing and proofreading, the more accurate the final translation and the more likely there will be a good working relationship with the client and repeat work.
MT is fine for personal use, especially when someone is browsing a website in a language they don’t understand or when they are traveling or communicating with a stranger. It’s certainly not perfect, but perfection is not the issue. You can’t always wait for a translation to appear just when you want it and there’s not always a suitable bilingual person around to interpret something for you and anyway perfection isn’t always necessary every time.
With the trend towards the internationalization of business and the growing need to communicate across borders quickly and efficiently there is a huge incentive to improve machine translation technology. No one can predict just where the technology will be and how much it has replaced human translators in the near to medium-term future.