top of page
  • Writer's pictureSijin Xian

AI, Machine & Translation

Advancement in artificial intelligence and machine learning has sparked considerable curiosity and concern about the future of jobs. While this might seem like the newest topic in town, it is hardly news in my corner of the world. Long before everyone marveled at ChatGPT's ability to write poems and algorithms alike, language professionals have been grappling with the ilks of Google Translate and DeepL.

But something is shifting. Earlier this month, I attended a virtual conference held by the American Translators Association on Translating & Interpreting the Future: Empowering Professionals to Innovate and Thrive, which brought together over 500 language service colleagues from over 35 countries. The widespread interest speaks volumes about our collective recognition that the days when machine translation automatically equaled laughable mistakes have long gone and that talking about our future in the context of technology is a conversation worth having.

And for you, a client who commissions translation services, it's equally important to understand the implications of these technological developments.

Technology is cool.

I love translation technology. I love that American tourists can get around in Italy with quick language assistance on their phones. I love how my CAT tools store my translation memory and help me stay consistent. I love being able to type less and create more in my MT-equipped audiovisual translation workflow. It's incredible how far translation technology has come, and it's exciting that it will only get better.

However, while translation technology has become powerful and convenient, it does not render human linguists obsolete. Not every text lends itself well to machine rendition. This varies on the formulaic-creative spectrum because the translation does not always mirror the original word-for-word. The more you detach from the wording to factor in linguistic and cultural particularities and to bring out the deeper meaning for resonance, the more challenging it is for the machine to replicate that process.

But on the flip side, this is not a reason to dismiss technology. The machine is learning very quickly and has a lot of potential. It's critical for language professionals to always be in the know and think about how technology can be used in the service of their work and their clients instead of outright rejecting them as the enemy.

Cool can be deceiving.

Have you read ​this New York Times story​ about a lawyer relying on ChatGPT to prepare a court filing and ending up submitting a list of relevant cases that do not exist? Examples like this show that among technology's many powers, the appearance of correctness is also one of them. The same thing can also happen with machine and AI-powered translation. Like any tool, it must go with capable hands and discerning eyes.

Additionally, before you start feeding your text into the mouth of the machine, double-check to see whether there's any sensitive content that should not go into the world wide web. Machine translation services may store or analyze the data, potentially leading to breaches or unauthorized access. It is essential to consider the privacy policies and security measures of translation tools or service providers to protect sensitive information.

Another under-discussed flaw of the machine is that it processes information indiscriminately, lacking the discernment to question your input, which is your source text. On the contrary, translators give your text an ultra-close reading, allowing us to spot easy-to-miss errors that might hinder the reader’s experience. To deliver your message effectively, we also serve as the reader’s surrogate in gauging the clarity of your text. While not hired to fact-check, we inadvertently notice incongruities when researching unfamiliar or technical terms and concepts.

Your translator needs to be cooler than technology.

The future of translation lies in expertly calibrated collaboration between humans and machines. If your translator only translates on the level of the machine, they might not pick up where the machine lacks and even get misled by the generated output. Conversely, you might also miss out on opportunities to be more efficient if your translator insists on eschewing all technology.

You need someone who has a masterful command of the languages, subject matter, and technology so that they can be a capable and responsible shepherd and gatekeeper of the translation process. This will allow for the best of both worlds, combining human translators' expertise and creativity with AI's efficiency and convenience.

In conclusion, while the advancement of artificial intelligence and machine learning has transformed the translation landscape, it does not mean the translation profession is dying. Technology has its merits but also its limitations and risks. As a client, it's in your interest to be mindful of privacy concerns, the potential for errors, and the need for human expertise to ensure accurate and effective translations.

Hi, I'm Sijin! I'm an ATA-certified English<>Chinese translator dedicated to getting things right and doing things right. Are you wondering if machine translation might work well with your texts or looking for someone to ensure the quality of your translation output? Get in touch today, and I look forward to helping you get the best of both worlds.


bottom of page