Machine Meets Human
Machine translation is carried out entirely by computer programs. These babies have the advantage of being quick and easy to get your content translated. However, they are often extremely inaccurate! Computers, unlike humans, are not able to create meaning from a combination of words they have never seen before.
But we get it, we get it! The rapid pace of improvement of machine translation technology makes us think a bit more about the future of translation. After all, there is artificial intelligence, Word Lens programme that recognizes text through a mobile phone camera and translates it simultaneously into 27 languages and then, Google Translate’s conversation function available in 32 languages. It can’t be so bad! We can, in fact, see how this technology could enable a company to speed up tasks and save costs.
Well, talking about the future and machines taking over, we can chill for now. First of all, stop worrying. If there’s any way the advanced AI enters this field for good, is to help us, not to steal our jobs 😉
But the usual results of using just Google Translate, are pretty often incorrect. Sometimes it’s even impossible to understand what the context of the text actually is. After all, professional translators and linguists mocked Google Translate for many years… not without a reason:
What Did The CAT Do?
Now, computer-assisted translation is not nearly similar to machine translation. Thank God! Here, the translation is carried by human translators with the assistance of translation software (such as Text United). That increases the efficiency of the human translation without the loss of quality.
If you want even more automation, there is Translation Memory. Every time a translation unit reappears in the text, once translated, the identical text will be also translated throughout the whole file and for similar sentences. You can use translation memory proposals and adjust them. Whether you are translating yourself or ordering a translation, translation memory saves translated segments. All translation memories which include segments in the same language combination as your project will be linked to the newly created project.
Human Meets Machine
The results of using machine translation are getting better, as of right now they are often useful for getting the gist of a text. But remember that most of the time machine translations come back awkward if that could be the only problem. Those are not good odds, and incorrect grammar or not-quite-right word choices can obscure meaning.
When it comes to human translation vs machine translation, the choice is obviously the best of both systems. To get the both, choose the platform that will look up your translation memory for any previously translated similar sentences and for untranslated segments, with no matches in the translation memory. The translator can insert the machine translation there with a single click (and edit, if needed).
This method will give you all the advantages of machine translation but you will be able to edit them right away and save making your translation even more automated. You can turn on automatic machine translation, so after opening each segment, the sentences are there thanks to machine translation-ready to edit. I just have to mention that at Text United, we have it all!