Technology has come a long way since the days of box Apple computers and floppy discs. With each step of development (called tech disruptions), we achieve better skills at shaping technologies into something increasingly more accessible and useful. Now in modern times, the internet stands tall as the one of the indisputable achievements of human innovation. Initially a tech disruption itself, the internet has become so widely adopted that it now also experiences its own periodic evolutionary changes, such as Web 0.0 - Web 5.0. In each upgrade, users gain greater control of the many systems that make it up and then shape and use it as desired.
Along with the rapid growth of the internet and its many components came the development of translation technologies. Now with the internet becoming a means to easily connect people and translation becoming easier to automate, everyone from large organizations to individuals at home can reach out and communicate with others across the globe.
In light of this, we're taking a quick look back at the growth of the internet and where technology is headed, both on the web and in language translation technologies.
The S-curve (the 'Sigmoid function' in mathematical terms) is a data scale expressing how there is an expansion of scalability. Essentially, what was once accessible to a few becomes, over time, accessible to many. The S-curve in this case represents proprietary services, tools, utilities and features. Just as when building a house, once the foundations are laid, you build upon them, adding beneficial features and functions while continuing to further improve upon them.
Just as how the web developed and became simultaneously more intelligent and accessible, the S-curve is important to consider in the development of technologies and software, as is it a requirement of evolution for systems to expand and grow. By following this model, the transition of machine translation (MT) to computer-aided translation (CAT), to artificial intelligence (AI), then to quantum computing (QC), is seemingly inevitable.
Machine translation, or MT, still has a lot to offer. A great deal of MT is the execution of backend coding. This is particularly accurate when it comes to finances, like exchanging currencies.
Paired with AI, MT isn’t going anywhere with our current coding system, as it’s already been fully integrated. With Web 5.0 we will see the introduction of a new system format that will begin to integrate the new AI QC systems, which will not only eliminate MT but the way we fundamentally interact with codes and processes. It will be future tech disruptions (like the far-flung Web 6.0-7.0) where we will really see a shift in technology, where hardware and software will be phased out and replaced entirely with AI/QC-compatible software.
In our current web rendition, MT on the front end has one major drawback in translation - it lacks the ability to discern culture, locale and social differences, and therefore often requires substantial human intervention in the form of post-editing. Trying to get a correct raw translation from open-source MT engines, like Google Translate, can be problematic in certain languages and content types, as the system is set to translate via dictionary and not social dialect. This means mis-translations can be a common problem. At its base, MT is not intuitive, and that is quickly becoming an issue as intuitive AI begins to show up in more systems.
Computer-aided translation, or CAT, is the process in which human translators use computer software and hardware to support and facilitate translation. At this time, it is the best guarantee to clear and concise translation efforts. CAT processes can include writing systems specific to regional (locale) translation with document editors, terminology management and translation memory.
CAT is probably the one system in computer automation that keeps humans in the daily process. With CAT, the need for human eyes to review work is still an integral part, making it a fundamentally human effort and one that offers the highest potential for context success or flaws. Where MT is not intuitive and cannot translate with precision, human review can sense the nature of the work and change it appropriately, ideally to yield correct and natural-sounding lingo.
We will likely see an end to CAT systems in the next decade with the advancement of AI as humans have the potential to suffer interest of misguidance, misinterpretation or bias towards content. Theoretically, given AI’s complete lack of social interest or emotional bias, it will be able to translate within the correct locale without any inserted or unintentional bias. Furthermore, with QC, AI will be able to measure the tone and emotional reception of words and apply translations within the locale context without misinterpretation.
At Andovar, we are as interested as you are in the evolutionary steps of the internet and the world at large. We want to be right there at the forefront of technology disruptions, integrating new system updates as they become available. With that in mind, our staff has practiced long and hard to become experts in the best systems presently available, and right at the front in this age is CAT.
Our team is prepared to help you access the best translation hardware and software available with the promise that as soon as new tech is available, we will be there to help you smoothly transition to them in the future. Andovar’s Language Technology Tools are here to make your life a little easier.
Please feel free to get in touch if you have any questions about translation technologies or to see how we can help you with your next localization project!
Find out more about website translation technologies in our free Website Translation Ultimate Guide.