Machine Learning is a big scary term for something that can be explained very simply. When you see someone you’ve seen several times before, you recognize them. In the same way, Google’s computers recognize photos of your face because they all look similar. Google’s machines are able to learn – that’s just one piece of the puzzle. Google has announced a system called TensorFlow, a machine learning system that Google will be using soon (in place of their old system being used today), and they’ll be making it open source, too.
TensorFlow will be in place at Google ASAP on all systems where machine learning is best put to use. This includes implementations like photo identification in Google Photos, Smart Reply in Inbox by Gmail, and speech recognition in the Google search app.
Below you’ll see how Google is using machine learning and will be using it with TensorFlow through the future.
This system is extremely scalable, being able to be implemented and used on devices as small as a smartphone and as large as a datacenter of full-scale computers.
According to Google‘s CEO Sundar Pichai, this new TensorFlow “allows us to build and train neural nets up to five times faster than our first-generation system.”
Google has decided to not only use this system on their own software, but to release it to the public as well. Google is making TensorFlow entirely open-source.
“We hope this will let the machine learning community—everyone from academic researchers, to engineers, to hobbyists—exchange ideas much more quickly, through working code rather than just research papers,” said Pichai, “And that, in turn, will accelerate research on machine learning, in the end making technology work better for everyone.”
You can learn more about TensorFlow – and use the smart system for yourself – at TensorFlow.org.