The Impact of AI on the Quality of Language Translation

Business Owners, B2B, B2C

Introduction

AI took the world by storm over the last few years. It wasn’t invented recently, but what I mean is that it has been a hot topic of discussion in every career field including the language industry.

Success is cultivated! It may be, in your case as a language professional, that you need to keep up with trending technologies. Language evolves and so does technology. In this article, I will focus on the advantage of human language service providers over AI.

How it all started

In 1935, a British logician discovered a computing machine that would be able to do more by reading and writing thanks to its sophisticated memory. The evolution of this theory took several years but finally flourished in 1950’s, causing giant computer technologies to embrace it.

The area of language processing known as computational linguistics was not affected by this technology until 1950’s. In this area, AI emerged as a machine translation process through which the United States military department started to automatically generate translation of Russian text into English.

With the Coronavirus pandemic threatening the economy and limiting the ability of the global workforce to operate in a vigorously robust way, companies started to invest in basic long-lasting solutions that can be assisted by humans. Until now, businesses are racing to secure AI in place for assured convenience and, in some cases, to cut financial costs.

The language processing systems have allowed the creation of automated processes in the technology field. Speech recognition is one of these areas. Under computational linguistics, AI has now allowed for the development of technologies that enable the recognition and translation of human language into text form. This is completely done by computers and other technological devices.

Human effort remains central

The language industry has enjoyed a variety of services over the years. It is unarguable that the human role is central to these services. To be clear, even productivity-enhancing technologies such as the development of AI totally require the human brain. The same is true for language services. These are services that cannot be replaced.

One of my major clients who has embraced process automation has recently added a layer of disclaimer to their project kick off. Since the very first stages of their project development are carried out by machine, they seek feedback on the quality of the machine’s output. This step is followed by a series of quality checks by humans to ensure the best quality final product.

In the traditional human translation process, linguists would handle the process using machines, software, or tools. Since the evolution of AI, machines are able to handle complex processes using preset computational rules. However, language experts have the final word with regard to the accuracy of the machine’s output.

To what extent is human effort affected?

One of the industries that are leveraging AI is e-commerce. In their research about the effect of AI on international trade published in December 2018, the Washington University in St. Louis demonstrated that an AI tool had boosted eBay’s quality in listing title translation by 10.9%.

On the future of work, McKinsey Global Institute discusses that automation will likely displace around 15 percent of the global workforce, or about 400 million workers in the period 2016–2030. These figures may also double, which entails that 30 percent, or 800 million workers may be displaced by 2030.

While 77% of people express their apprehension that AI could bring about job losses in the next year, some other areas will likely need more workers to operate the technologies created by this trend. The gap will be filled in two ways. Some areas of operation will need to reskill while others will augment their human resources.

AI limitations pose a staggering concern

There is a growing concern among end users about the accuracy of the technologies disseminated by AI across industries. This is especially the case on the internet where search engines have been overpopulated by automated content. A research shows that over 75% of consumers are concerned about misinformation from AI.

Of the free tools available for instant use, some machine translation tools such as Google Translator, for example, present staggering results which can put users’ projects on the line if completely relied upon. I find this to be the case for all machine translation tools.

Some processes cannot be replaced by automation

About five years ago, I was hired by a local LSP and tasked with coordinating translation projects for some of the leading automobile brands here in North America which included Honda, Mercedes, Polaris, and Arctic Cat to mention a few.

These projects were so meticulous that I spent a considerable amount of time on each of the eight steps each went through to make sure that every phrase was absolutely accurate. Plus, there were additional review and QA steps that were carried out by different team members. The team effort helped us to achieve desirable results.

This is an example of a linguistic service that cannot be replaced by AI. Nonetheless, companies have decided to dedicate a chunk of preliminary steps to machine translation so that subsequent steps can be carried out by humans. This is where post-editing kicks in.

It is important to note that MT boosts speed by multiplying project timescales, but the end result is achieved by the efforts of a human being. Some of the industry-leading language service providers discourage machine translation so much that they disable it from their software. Its benefits are indeed remarkable but high control of the results is required.

Even though positions such as project management, editing, and quality assurance cannot be replaced by automation, a staggering function will decrease drastically so that human resources do not have to do as much. This scenario will affect both the workforce size and the length of the process.

What businesses should watch out for while using AI

Artificial intelligence has been such a nice tool for people who want to look up things on the internet, find their way around places, and sometimes invest in some basic self-development activities such as language learning. Can companies benefit from this process?

When it comes to doing business, this is a game we all must be serious about. Given the concerns I elaborated on above, no one wants their business to be discredited due to the use of misinformed data found on the internet. To be successful, information must be collected, researched, validated, and launched so that consumers have reliable data.

This cannot be achieved by leveraging AI. As I mentioned above, there are some basic benefits to using open-source tools such as machine translation. The evolution of technology is obvious and unquestionable.

If you are handling large projects which you wanted to tackle all at once, it would be worthwhile to use machine translation, for example. In this case, you must ensure that you have enough financial and human resources to fix discrepancies that will result from the use of AI.

Conclusion

AI should be used to tackle l large-scale language projects. Since the process is time and finance consuming, there are a lot of things at play before you decide whether you should use AI-powered processes such as project automation, machine translation, or computer-assisted translation tools.

The project automation has already been out there for decades but has always required skilled and knowledgeable human resources to produce desired results. I gave you an example of machine translation post editing above, which requires expert reviewers and proofreaders to make sure that final versions of your documents are accurate.

To successfully showcase your products in different languages, you must have full control over their translation and localization. It is important to remember that businesses who want to be successful and maintain the quality of their products and services must embrace human efforts over AI.

Message to Our Readers:

We hope you enjoyed this article. Our goal is to add value to our clients by making their businesses profitable and hassle-free. If you would like to learn more about how our language services can support your specific business model, we would be pleased to connect and discuss your questions further.

Click here to contact Sim directly. If you are a Language Professional aspiring to success, click here to check out Sim’s new book— “Essentials of Career Management for Language Professionals”. Watch out for our next article in the next month.