Google announced an entirely new machine learning system known as TensorFlow. The new system can be adapted easily to new research, since it is more faster and flexible than the old system.
As stated in the official blog by Sundar Pichai, CEO, Google,
It’s a highly scalable machine learning system—it can run on a single smartphone or across thousands of computers in datacenters.
Stating further he said that we can use TensorFlow for speech recognition in the Google app, Smart Reply in Inbox, and for search in Google Photos. Surprisingly, it can be used to improve our products quickly with a speed five times faster than our first-generation system.
Also, today they’ve announced that the TensorFlow will be open sourced. So, the machine language can now be worked out by many people including engineers, academic researchers, etc.