A February 2018 survey by linguistics and website localization agency Locaria found that many professional linguists worldwide had concerns about AI’s impact on their work. Some four in 10 respondents said they worry there will be less work for them, and 34% expected they’d have the same volume of work, but with expectations for higher standards.
MT technology has improved dramatically in recent years, enabling consumers to understand web content in different languages and helping businesses do faster, higher-quality and lower-cost translations. Other types of AI, including machine learning, are also streamlining translation and localization project management, saving time and resources for linguists and marketers. A growing number of language services companies now use sophisticated software systems to identify the best translators for a job, farm them out to linguists and send them back to clients.
But while AI-driven machines often excel at repetitive work and word-by-word or sentence-by-sentence translations, they frequently struggle with the linguistic and cultural nuances of localization. Marketing copy is even harder for machines to get right, and really needs a human to nail the right tone of voice or a specific geographic region’s cultural preferences. This is where linguists and other language service professionals can still add value: A human touch is still needed to tweak machine translations so they sound more human and natural.
Technology is providing many opportunities for these professionals to optimize and streamline their workflows. The traditional method of translating, for example, would require a human translator to transcribe audio and then translate, which might take hours. But those who work with MT technology can run a file through a translation program, get a rough machine translation and then fix errors and make the language sound more human, which takes less time. This removes some of the more mundane work from a translator’s process and allows them to work more efficiently.
Despite these potential benefits, the Locaria study found that 73% of linguists said they never use statistical machine translation (SMT) in their work, and only 16.4% said they regularly use it. The same linguists also said they were less aware of AI and NLP technology than expected, with 38.2% saying they had no knowledge at all, and just half saying they had heard of AI.
Most industry experts believe it will only be a matter of time before humans and machines work symbiotically to produce the best work. “These new technologies are having a significant impact on the language services industry,” said eMarketer principal analyst Victoria Petrock. “And while they still can’t replace human expertise, they are dramatically changing the workflow. People who embrace the technology and use it to their advantage will ultimately come out on top.”
Courtesy of eMarketer