In a perfect world we would all speak the same language and therefore we would be able to express ourselves and understand the different cultures of the world much more easily. Unfortunately, this perfect world does not exist, and each country has its own peculiarity.
Still, it’s a difference that we can celebrate, but that doesn’t help break down the cultural barriers that separate us. But that’s the job of Google Translate, Google’s already very powerful tool for translating a text, a photo, or even a live chat into 103 different languages.
This tool is being updated today to finally include what the developer presented to us at the Made by Google conference: Neural Machine Translation. Using his first steps in artificial intelligence, he was able to make Google Trad even more efficient.
This process is currently only 8 different languages : English, French, German, Spanish, Portuguese, Chinese, Japanese, Korean and Turkish. Eventually, it will be developed for the 103 languages the tool recognizes.
What change does it bring? A more human translations quite simply being able to recognize a text and translate it in a much less robotic but natural way into the language used. Google shows in its example that Neural Machine Translation has greatly improved machine translation.
Since the launch of Assistant, we think the developer has been keen to use machine learning in all aspects to make their products more efficient. Recently, it’s Google Play Music’s turn to grow out of this mill. Hopefully these efforts will bear fruit.