Will we still need human translators?
Language lovers and language professionals alike are keeping a nervous eye on developments in machine translation technology, watching for the day when a computer application such as Google Translate, Bing, Moses or Systran delivers a translation that is just as good as what a human can produce.
Machine translation vs human translators:
Can machine translation replace a human translator today?
No, not if perfect quality is the goal.
One day, the so-called deep learning engines may be able to perfectly decode then re-construct all of human communication. Until then, however, the interesting question is not whether machine translation (MT) can replace human translation (HT), but how human and machine can work together.
If Lexcelera has been a leader in MT for corporate use, that’s because we see technology as an enhancement to human skills, not as a replacement, allowing us to offer value added services that are built on the intersection of human talents and trusted technology.
The questions that drive us forward are:
- How can MT contribute to the work of human translators?
- How can human translators contribute to the quality improvements needed by MT?
One of the ways machine translation contributes to the work of translators is as a productivity enhancer. To produce translations of a professional quality level, translators can complete their work in around half the time by using MT. How? It’s easy: the MT engine generates a first draft translation that they then post-edit.
The fact that this productivity gain does not hold true if the MT engine is poorly trained brings us back in a virtuous circle to the human talents behind all good translation automation.
Machine translation vs human translators: the irony
Without human translators there is no machine translation
- Statistical (SMT) engines are constructed from millions of segments of human translations;
- Rules-based (RBMT) engines have at their base grammatical rules and bilingual dictionaries built by – you guessed it – human linguists.
We predict that MT is here to stay because of its usefulness to both translators and customers alike.
But far from seeing machine translation as an imminent threat to the livelihoods of translators, we see translation technology as intertwined with human linguistic capabilities.
For Lexcelera, connecting talents to technology is where the magic happens.