It’s fair to say that computers form part of most people’s lives and we rely tremendously on technology and software to run our businesses and lives in general. Whether our chosen profession is accountancy, architecture, broadcasting, media, education or medicine (to name just a few), it’s very difficult to imagine how we would deliver good quality products efficiently without the use of computers.
International culture has now passed through and enormous transition period where we all converted from using pen and paper to digital communication. However, communication is complicated – but is it too complicated for artificial intelligence?
When it comes to translation in the form of writing and speaking, the big question is could computers do the job of human translators? Absolutely not. Translation is communication in its rawest form and although artificial intelligence sometimes is mind blowing, it’s not sensitive enough to replace the brilliance and the human mind.
1. It’s Difficult to Create a Top-Quality Translation
Firstly, it is very difficult to create a top-quality translation even if a human is a bilingual perfectionist. Translation requires more than just a word for word recognition and good working knowledge of two languages. To get it right you must recreate and craft word combinations, grammar and sentences so that a translation has the reflective impact of the original text or speech. This is no mean feat and is something that a computer cannot successfully articulate.
2. Human Translators Produce Differing Language Conversions
Secondly, humans are all different by nature and therefore human translators will always produce differing language conversions. There are many factors that affect their work such as emotional context and expression. Given this level of complexity you may receive many different versions of a translation if you were to ask a number of qualified and competent translators to perform the same piece of work. Although computers can learn, it’s thought to be near impossible for computers to generate such an understanding of human emotions as there are too many variables. As a result, a computer generated translation may appear flat and lifeless thus losing the intended message.
3. The Volume of Languages is Simply Too Large
Thirdly, there are an estimated 7000 languages in use across the globe. If you consider the combinations that are possible, the sheer volume is enormous and something that would take hundreds of years to develop and test successfully.
4. Computers Can’t Understand Context
Lastly and most importantly, the context in which a word is used is incredibly important. Let’s take the word ‘run’ as an example. Peter Gilliver, an Oxford Dictionary lexicographer states that this single word has 645 distinct meanings. It’s doubtful that any computer could learn these meanings and translate to another language accordingly. However, context is embedded into the human mind.
In conclusion, computers and software are indisputably useful to us and although we could use machine translation tools to help us, it’s very unlikely that they could never replace human translators.