Google Translate has been around for over 10 years and if you use it frequently you may have noticed that the service recently received an update that significantly improved its accuracy. This is thanks to the use of neural machine translation, Google used the latest AI advances to help its translate service to better understand language. Thanks to machine learning, texts converted to a foreign language via Google Translate now have better grammar and read a lot more natural.
To help you understand what's going on behind the scenes, NY Times published a massive long-read about how Google used AI in this instance, and how machine learning is reinventing computing itself. You can read the piece over here: The Great AI Awakening.
The paper notes that the effort took just nine months to complete, which is a lot faster than Google's own engineers expected. By adopting machine learning, the Google Translate service booked as much progress as it did over the past ten years combined.
At the moment, the improved version of Google Translate is exclusively available for English, French, German, Spanish, Portuguese, Chinese, Japanese, Korean and Turkish. Google plans to roll out the new AI-base system to abut eight new languages each month, meaning the entire effort should be completed by the end of 2017.