AI meets human nous: the future of the translation industry

The future: it’s big, it’s unwritten, and we’re all heading towards it. And while many aspects may remain unclear, the technological developments of the past few years—the rise of ChatGPT, big data, machine learning and the like—indicate that artificial intelligence will play a key role in how we function as individuals and as a society. AI is shaping the future of the translation industry, too, and with 2024 on the horizon, now is a good time to take a look at how our field of expertise could flourish through the integration of AI-supported solutions in the coming years—as well as a few red flags companies should keep in mind.

Artificial intelligence: a double-sided digital sword

Remember last November when ChatGPT took the world by storm? Here, beamed to us from the future, was an AI tool that could create texts on everything from mountain biking to metaphysics in a matter of seconds. According to the developers, the more people used it, the better the AI would be trained and the more precise the texts would become. Soon enough, doom-mongers were ringing the death knell for copywriters, translators and marketing experts everywhere.

There were just three tiny problems: first, the texts produced by ChatGPT, while serviceable, are rather generic. Flair, wordplay, poetic license, style: this is not what an AI is programmed to provide (not yet anyway). Second, the content lacks the human touch, both on a linguistic and substantive level; after all, personal insights and a love of one’s craft are what make a text stand out and keep readers engaged. Third, not everything the AI reproduces is correct. That isn’t to say the tool is being purposefully dishonest, but there are instances of it regurgitating false information it finds online. Therefore, as we embrace AI tools for what they can do well—kicking your brain into gear with initial ideas and creating a text outline—it will remain essential to have human beings on board who are experts in shaping texts to sound human, engage readers and be truthful.

Post-editing for a post-everything world

So you’ve discovered that DeepL and Google Translate are now ‘good’ enough to translate your memos, boilerplate texts and T&Cs. But wait, you think. I’m not a linguist. How can I know if that English idiom reflects the German one or if ‘recalcitrant’ is really the best word in this context? The answer: you can’t, but your trusted language service provider can. Only a language expert can take a machine-translated text and properly assess it for its accuracy and correctness and make changes that enhance rather than detract from the overall result. This is what the industry calls post-editing, and it will likely constitute the bulk of our work in the future. By all means embrace these innovative technologies if the y help you save time and money, but be sure to keep a good LSP on hand to scrutinise and tweak the results before you publish them.

Transparency is the key to any good working relationship

We aren’t naming names, but we have begun to observe a vaguely sinister new trend: certain LSPs are playing fast and loose with the truth when it comes to the translations they provide. They claim the texts are translated by a human, but even a brief glance reveals mechanical sentence structures, uneven pacing, inconsistent terminology and a detached tone. In short: if it sounds like a machine, then it was translated by a machine. As a long-term strategy for success, this is a very poor one. The uncanny valley effect means customers will quickly realise the text hasn’t been edited by a human.

Now, there is nothing wrong with machine translation in itself (we at SwissGlobal offer it as well): in an age of rising costs and ever-shorter time windows, it is an excellent solution for translating certain documents. But as we race headlong into a future defined by data, artifice and technology, it will become more important than ever before in the translation industry to guarantee transparency surrounding the types of services provided. That means not simply running content through a machine, but fiercely defending the human touch as an integral part of the process.

Privacy and ethics are yours to claim

Are you aware that any information you paste into an online tool will be stored on a server somewhere in the world? Or that your data is used to train future iterations of AI solutions like ChatGPT? As we have pointed out time and again, it is essential to protect your data, not least from hackers and the competition. The digital world is sadly a predatory place, and if you are using online tools without prioritising your privacy, this could have severe ramifications for your business.

The ethics of using AI is another aspect we will need to factor into business decisions going forward. This is because AI-supported results are subject to bias, inaccuracies and discriminatory outcomes, which can have negative long-term effects on how we human beings perceive the world. Thankfully, as an article in Forbes points out, 2023 has been a year for more legislation governing ethics in AI. That means creating frameworks to mitigate the risk of propagating harmful or false information and setting audit requirements regarding data bias, privacy and security. In other words, now is the best time to put privacy and ethics at the forefront of your company’s policy on AI use, because tomorrow may be too late.

Navigating a new era of translation safely

The future of the translation industry is full of innovations that will change how we create texts—but it isn’t without its pitfalls. Think you’re ready to enter this brave new world of translation, but have reservations about technology, transparency and trends? SwissGlobal is here to help. As Switzerland’s most secure language provider, we fully understand the risks associated with AI and machine translation when they aren’t in the right hands. We know how to get the most out of these technologies while delivering quality results you can always rely on. Get in touch today to find out how to speak the language of the future fluently.