Following decades of development, machine translators now use technology inspired by the human brain. This was so in a competition at Sejong University in South Korea in 2017, when four professional translators were put up against three machine translators.

Each of the machine translators in the competition had, the year before, incorporated new translation mechanisms that used neural networks. They are human brain-inspired artificial intelligence systems that analysed the context of sentences and chose the more suitable translation after having examined an extensive database of translated materials.

The machines won on speed, but it was quality that was being tested. In this respect, the recent achievements of machine translators were not enough for them to win out against the human translators in Sejong. All three machine translators finished behind the professionals. This time the problem was not the context, but the machines’ inability to analyse their own work. The International Association of Professional Translators and Interpreters awarded the human translators 25 points out of a possible 30, while the best machine translator was awarded only 15 points.

Event organiser Kang Dae-young believes text translators are an increasingly useful tool for humans: “To adapt to the new age of machine translators, we have to learn to work with the many available tools. The role of the human translator of the future will be to complete the final machine translation.”

Leave a Reply

Your email address will not be published. Required fields are marked *