Artificial intelligence and data will definitely be the new oil in the second half of the 20th century. So, have the big tech companies struck gold in terms of rendering translators obsolete? Not so fast, some would say. Tech writers have called it a misconception of the highest order to assume that deep learning coupled with more data will automatically result in neural nets displacing humans in general and translators in particular.
You could well be thinking that all of this sounds great but doesn’t have real-world evidence to back it up. As it happens, these claims have some basis in recent research. A translation team from South Korea pitted a crack team of translators against the most innovative artificial intelligence around. The results demonstrated that machine learning has some ways to go. An estimated 90 percent of the translated material produced by machine learning was said to be grammatically clumsy.
In fact, artificial intelligence has a long way to go in terms of producing translation results more efficiently than its human counterparts. Artificial intelligence might be an efficient way to quickly translate a bunch of material to get the gist of what the original translation is attempted to say, but machine learning still produces results that fall below what a native speaker of the source language meant.
How America’s Top Tech Companies Are Approaching AI Translation
With that said, how are America’s top tech companies approaching this issue? Well, Google has a handy device called Google Buds. How do they stack up? Google Buds are basically wireless earbuds that allow users the opportunity to get a quick translation of a particular text or podcast. The experts, though, have found that Google Buds deliver awkward results; the results are so awkward, in fact, that Google Buds would be inappropriate for most real-world contexts. For more high-stakes translation services, such as conference translation services, forget Google Buds and go with a team of human translators who all understand nuance and subtle syntactical issues between languages.
Where does this leave us? The fact is that the future is undoubtedly going to feature a blend of artificial intelligence and human ingenuity, and the translation industry will certainly not be an exception to this general rule. Artificial intelligence will be continued to be used to deliver serviceable translations in certain contexts. In some contexts, this trade-off will be acceptable – e.g., someone on the go who simply needs a quick translation of an email or podcast. The stakes aren’t life-or-death. With machine learning and artificial intelligence there’s going to be a trade-off between efficiency and nuance for at least the next decade.
The Future for AI Translation is Uncertain
More than a decade out, the future is more uncertain. With the help of human translators, machine learning could enhance itself further with the aid of human translators who can correct blind spots that would never be seen by a computer algorithm. In the short term, however, human translators can spot nuances that artificial intelligence routinely overlooks. For high-stakes translations, there’s no contest: Human translators and interpreters aren’t going anywhere soon!