MT is Changing the Industry, Just Not in the Way Mainstream Media Thinks It Will
Machine translation (MT) has caught the public eye once again. The Wall Street Journal recently predicted that “the language barrier is about to fall” – within 10 years. We might note that this is one barrier that has been “about to fall” for far longer than the actual Berlin Wall stayed up, and that it’s been an awfully long 10 years since the first such claims were made in the 1950s.
More recently an in-ear machine translation tool aimed at dialog – which may or may not be a fraud – blew through its funding goals to gain crowdfunded support of over $2,000,000 along with more credulous – and in some cases completely uninformed – press coverage than any actual, functioning language technology has in recent memory.
Unfortunately, we are in the midst of an irrational euphoria about machine translation – one that some sources within the language industry have helped stoke – and separating fact from science fiction can be tough. CSA Research has studied this market segment for over a decade, with an emphasis on gathering unbiased, primary data.
Since January 2016 we surveyed 900 global enterprises, LSPs, and freelance translators on precisely how they use MT in their daily operations. Two reports address the enterprise and LSP sides of the equation, and three briefs cover how increasing translation volumes affect content creators and language companies, and how LSPs can decide whether or not to offer post-editing services. We plan additional publications in the coming months that will further flesh out various aspects of the picture our research paints.
So what do we discover? The following findings inject some realism into the picture:
- MT is here to stay. Usage rates for MT have made steady gains in recent years, and we believe that it is near a tipping point: It is now on a path to become a mainstream solution for both enterprises and LSPs within the next three years.
- MT is not doing away with human translators. They have long worried that buyers will replace them with sub-par MT. Although some translators will lose some jobs to MT, overall demand for human translation will continue to grow.
- Post-edited MT is the growth opportunity for LSPs. Large enterprises expect double-digit annual growth rates in translation, growth that present methods cannot possibly keep up with, even if the language industry were to add new translators at a historically unprecedented rate. Demand for post-editing services will grow faster than any other segment of the language industry and will allow LSPs to maintain quality while simultaneously delivering higher volumes.
- New technologies are transforming the MT landscape. Post-editing is growing, but it no longer looks like what most translators are used to. They may have mixed feelings about the task, but new technologies such as LexTranslator, Lilt, MateCat, and SDL Language Cloud are changing how translators interact with MT. These tools each take different approaches to give control back to the translator while also gaining the productivity advantages that MT can provide.
Our research shows that MT is profoundly changing the landscape of the language industry. However, it will not eliminate the need for second languages or replace translators – certainly not any time soon and not within the ever-distant 10-year horizon beloved by tech journalists. The findings may not be as sexy as claims about how untested or experimental technology will up-end the whole way we interact across languages, but they present a much more realistic picture of the role that technology has to play.
About the Author