Google’s new machine learning framework to put more AI on your phone

APD NEWS

text

At the moment, artificial intelligence lives in the cloud, but Google — and other big tech companies — want it work directly on your devices, too. At Google I/O today, the search giant announced a new initiative to help its AI make this leap down to earth: a mobile-optimized version of its machine learning framework named TensorFlowLite.

The original TensorFlow was released in November 2015, and quickly became popular with both researchers and developers as a way to build AI tools. TensorFlow is flexible, reliable, and comes with a big stack of documentation that makes it easy for beginners to get started. The newly announced version, TensorFlowLite, will build on this, helping users slim down their machine learning algorithms to work on-device.

“It’s a library for apps, designed to be fast and small, but still enabling state-of-the-art techniques,” said Google’s Dave Burke. “We think these new capabilities will help power the next generation of on-device speech processing, visual search, augmented reality, and more.”

The company also announced that an API for making machine learning work better with phone chips would be coming sometime in the future — a clear sign that Google thinks your next phone will have an AI-optimized chip in it.

But why is it good to put AI on your device in the first place? Well, simply put, machine learning applications that run locally on your phone are faster, more private (the data never leaves your device), and don’t require a working internet connection. All this translates into a user-friendly experience that’s better for both customers and Google itself.

Other tech companies are trying to make this transition, too. Facebook has its own version of TensorFlowLite that it calls Caffe2Go, announced back in November last year. This has enabled the company to create neural network-powered art filters that work on users’ phones, transforming photos and videos on the go. TensorFlowLite should help Google (and the wider AI research community) bring even more interesting functions like this to our most-used and most-important devices.

(THE VERGE)