Translate is tapping into neural networks for smarter language learning

BY GreenBot Staff

Published 16 Nov 2016

Translate is rolling out a major upgrade that promises more human-like language translations.

is bullish on its Neural Machine Translation technology, claiming that it’s a bigger upgrade to the service than everything that’s been accomplished in the last ten years combined.

The company is rolling out the improvements to eight language pairs in search, the Translate apps, the website. You’ll find the new technology behind translations between English French, German, Spanish, rtuguese, Chinese, panese, Korean Turkish. says that makes up more than 35 percent of all language queries.

Translate oduct ad Barak Turovsky said the new system looks at entire sentences instead of just translating one word at a time. Often times words don’t have a direct translation from one language to another, so this new approach gives a result that’s more akin to how humans actually communicate.

Since it’s easier to underst each sentence, translated paragraphs articles are a lot smoother easier to read. And this is all possible because of end-to-end learning system built on Neural Machine Translation, which basically means that the system learns over time to create better, more natural translations.

’s cloud intelligence isn’t just for the company’s consumer services. It’s a major component of its enterprise effort by infusing machine learning into G Suite other services.

To check this out, you can head to Translate online or grab the latest app for Android iOS.

y this matters: A better understing of human language is a critical lane in the race to build smarter artificial intelligence. Microsoft has also invested heavily in this area, most notably with Cortana its purchase of SwiftKey.