Media: Lianhe Zaobao
**Does Artificial Intelligence Have Little Impact on Translation? Really?**
I wrote earlier on Zaobao about the anticipation of self-driving cars. A friend mentioned that despite the complexity of autonomous vehicles—considering factors like the integration of technology into the vehicle, urban planning, regulations, testing, and a significant trust crisis regarding safety—it’s not easy to replace humans. Yet, in the U.S. and China, we see the rapid growth of autonomous vehicles. On the other hand, some jobs that seem easier to replace with large language models of generative AI, like editing and translation, haven’t been fully eliminated even after two or three years. Why is that?
If AI is so powerful, why is there still a high demand for translation work? Recently, when the language learning app Duolingo announced layoffs and replaced some employees with AI, the news quickly caused market concerns. The most notable targets of these layoffs were translators, an industry long considered one of the most likely to be replaced by AI. When this news broke, it seemed to confirm that AI’s impact on the job market had already arrived.
However, Duolingo’s CEO, Luis von Ahn, clarified this in an interview with *Planet Money*. He stated that the layoffs only involved 10% of their contract workers, not full-time employees, and that this decision was not solely due to the introduction of AI. Given Duolingo’s collaboration with OpenAI, the reaction to the company’s latest version of ChatGPT (GPT-4) is even more intriguing.
At the GPT-4 release event, OpenAI demonstrated its powerful real-time language translation capabilities. In a live demonstration, two employees had a conversation in Italian and English, respectively, with ChatGPT providing real-time translation. Although the demonstration was brief and successful, von Ahn pointed out, “Google Translate could do this eight years ago.” He further noted that computers’ ability to translate between major world languages is already quite mature. Indeed, over the past decade, AI technology has significantly improved machine translation capabilities, making it a key example of AI’s impact on the labor market. Even within the translation industry, AI’s application has not led to the anticipated large-scale wave of layoffs. So, why hasn’t AI completely replaced translation work?
Since Google launched Google Translate in 2006, the translation industry has been speculating on how AI will change the field. According to Bridget Hylak from the American Translators Association, the advent of neural machine translation (NMT) technology in 2016 brought a new technological revolution to the translation industry, with AI becoming part of the daily work of many translators and interpreters. Yet, global demand for translation remains strong. Data from the U.S. Bureau of Labor Statistics (BLS) shows that from 2008 to 2018, translation and interpreting jobs grew by 49.4%, largely due to globalization. Although BLS’s data collection methods changed after 2018, U.S. Census Bureau data indicates that from 2020 to 2023, the growth rate of interpreter and translator jobs still reached 11%.
In fact, despite the continuous advancement of AI technology, the translation and interpretation market has not been significantly impacted. Major companies and government agencies are still actively hiring translation talent. For example, Honda is hiring Japanese interpreters for its South Carolina plant, Starplus Energy needs several Korean interpreters for its Indiana plant, and the San Francisco City Government is hiring a bilingual (English-Spanish) translation proofreader.
According to BLS forecasts, the translation and interpretation industry will grow by about 4% over the next decade. While this growth rate is slower than the explosive growth of the past few decades, it is still slightly higher than the average growth rate of all U.S. jobs.
So, if AI’s translation capabilities are so strong, why are there still so many job opportunities in this industry? MIT economist Daron Acemoglu believes that AI’s capabilities have been exaggerated. He points out that although AI performs well in some areas, it is still not reliable enough for translation work.
Duolingo’s CEO von Ahn agrees with this view. He noted that although AI technology has improved translation efficiency, the company still relies on human translators to check and ensure the accuracy of translated content, especially in critical areas like user interfaces. Additionally, AI cannot consistently master the style and tone the company expects, so human translators are still indispensable in these aspects.
While AI technology is changing the translation industry, it is more of an auxiliary tool to improve translation efficiency rather than completely replacing human translators. As technology continues to advance, translators who can master AI tools will be able to find more opportunities in this market, while those who fail to adapt to technological changes may face greater challenges.
Overall, although AI technology has made significant progress in the field of translation, the future of the translation industry still requires human expertise to ensure accuracy and cultural adaptation as the final safeguard. As the data shows, demand in this industry is still growing, and the collaboration between AI and human translators will become an important component of the future translation market.
Some industries are developing more slowly but remain important and obvious, such as autonomous driving, unmanned logistics, and robotic restaurant services. People tend to focus on their own industries, often missing the bigger picture. However, it can only be said that the impact of artificial intelligence is inevitable. It could be a catastrophic disaster, but more likely, it’s just another auxiliary tool.
Talk to us to solve your digital challenges.
#AI #GenAI #ArtificialIntelligence #translation #chatgpt #gemini #google #tool #professional #digitaltransformation #industryinsights #duolingo #LLM #language
