We have been hearing for many years now that in the future, AI is going to feature in practically every area of business one can imagine. In many industries, AI is already playing an integral part in the activities of companies all over the world, whether it be in factories producing goods, warehouses sorting deliveries or in offices crunching data.
We have been hearing for many years now that in the future, AI is going to feature in practically every area of business one can imagine. In many industries, AI is already playing an integral part in the activities of companies all over the world, whether it be in factories producing goods, warehouses sorting deliveries or in offices crunching data. There has been a lot of talk about how the use of AI will go into overdrive in the future with the development of devices like self-driving vehicles and robots that clean offices.
The translation industry has already been greatly affected by the arrival of AI with automatic translation software and machine learning. It has never been easier to have translations from other languages available at the touch of a button, so does this mean that the millions of skilled professionals working in the translation industry are now surplus to requirements?
It appears there are many obstacles lying in the way of that happening. To understand this better, we have to look more closely into what machine learning actually is. What are the machines actually learning? They are learning sets of rules which are essentially determined by human programmers. An AI device will be able to improve its capabilities by collecting and processing data but this still depends purely on what has been programmed into it by the developers. This means it cannot think creatively or spontaneously, like a skilled translator can. Good translators know when rules need to be broken and exceptions to rules need to be made. So in other words, AI’s actions are simply repetitions of what humans have programmed into it. This is not necessarily a bad thing and positive outcomes can be obtained, but machine leaning must only be viewed from this perspective rather than be seen as the ultimate solution and a suitable outright replacement for translators.
The problem we have is that terms such as artificial intelligence and machine learning are being used by companies to give customers the impression that they are developing technologies far more advanced than they really are. We must understand that with AI comes both opportunities and limitations. It is important to not lose sight of its original intended purpose, which is to save time when carrying out operational tasks. What it can’t do is perform activities that humans perform spontaneously containing information that can only be obtained from a person’s own thoughts. These activities include writing personal emails, text messages, to-do lists and minutes for meetings. Then there are much larger activities such as whole translation projects, which require a wealth of background information of the task in hand including procedural exceptions and cultural considerations.
Machine learning however, is improving all the time and it is worth keeping an eye on its perpetual developments. Automatic translation is a useful tool today and its capabilities are only going to improve to the point where certain levels of context and nuance are understood. Despite these impressive developments, there are certain operations that a machine will always be unable to execute. These are the operations that rely on the finest of details to be correct in order to achieve the desired effect. Machines cannot change their functions in order to allow for exceptions to rules. In many cases, even top quality translators will disagree on the most suitable word to use in certain contentious situations and contexts.
What can be guaranteed however, is that machine learning will still play a huge part in translation projects, but not for actually doing the translations themselves. There are plenty of smaller tasks for which machine learning will come in handy. Simple message writing, glossary and memory building, data collection and indices are just a few that no longer require real people to do them. AI will not only take care of tasks like these, but will also be used in the administrative aspects of a project making everything quicker, more precise and helping to reduce costs.
When engaging in discussions such as these, it is important to understand what AI actually is and that its definition seems to be ever changing. Many tech companies will say they have already replaced humans with AI as they are using bots and algorithms for almost all of their operations. Others however believe that AI in its truest form hasn’t even arrived yet. This would be an appliance possessing the ability to incorporate creativity, cultural knowledge and moral reasoning into its operations just like humans do. It still remains unknown whether this level of AI is anything other than a fantasy. This doesn’t mean that tech companies are not going to attempt to find that out.
Right now and for the foreseeable future however, machine learning is still a way off from those ideas. What it is doing though, is enabling us all to become more productive. As mentioned earlier, for serious reputable translation projects, human translators will be required for many years to come until as some predict, humans will become a hybrid of man and machine, but that’s another argument altogether.
HI-COM is a localization agency dedicated to providing China-specific strategy, human translation services and certified translation to businesses and individuals around the world. Working with more than 100 brands, HI-COM is the go-to partner of companies that want to enter China market! Contact us for your free consultation today!
Scan QR code to follow Localization and The Chinese Market Entry News or Contact us on WeChat: