Google I/O 2017 Recap

By Adam Toennis on 31 May 2017

Google I/O 2017 has come and gone, giving us time to digest the keynote and technical sessions. The keynote announced a key strategic shift for Google, new improvements to existing products and even a surprise announcement!

Google shifts to AI-first

Artificial Intelligence is nothing new for Google and Sundar Pichai kicked the keynote off with an announcement a strategic shift from Mobile-first to AI-first. This change in direction was punctuated by the number of sessions covering, Google Lens and Machine Learning with TensorFlow. has a lofty goal to make state of the art AI accessible to everyone. Currently, the site feels a little sparse comprising inspirational quotes outlining their vision, some high-level links to their tools but most importantly links to blog posts. The site is worth monitoring as Google makes future updates to AI specific content.

Google Lens will let smartphone cameras process what they see in near real time and take action. For example, a user can point their camera towards a restaurant and retrieve reviews. Expect Google Lens to be integrated with Google Assistant

Sundar devoted time in both the opening and closing discussing Machine Learning both with AutoML and Tensor Flow. AutoML is a program that teaches other programs how to learn. According to Google, AutoML has already been used to design neural networks for speech and image recognition. While originally open sourced late in 2015, TensorFlow just had version 1.0 released earlier this year.

Updates to Android and Google Home

Google continues the rollout of Google Home by announcing it’s availability in five new countries: Canada, Australia, France, Germany and Japan. Google Home also has a new feature called Proactive Assistance. Proactive Assistance is essentially context-aware notifications on Google Home. When asked “What’s up?”, Google Home will tell you the notification (i.e. your normal route to work has heavy traffic).

Perhaps the largest announcement for Google Home was hands-free calling. Just ask Google Home to “call my mom” and it will be able to automatically detect the owner of the voice to provide additional context to the request and call the appropriate mom.

The upcoming Android O release will focus on two major themes: vitals and fluid experience. Vitals is described as Security Enhancements, OS Optimizations and Developer Tools. Fluid experience features are “picture in picture”, notification dots on the app icons and smart text selection. Google also announced TensorFlow Lite, a version of TensorFlow optimized for in-app machine learning on Android.


Google acquired Firebase, a mobile-centric “backend as a service” startup, in late 2014 to compliment its Google Cloud Platform. Firebase now provides services for mobile analytics, cloud messaging, authentication, cloud storage, remote config, dynamic links and much more. Firebase is a great platform to prove out an MVP for an early stage startup or for anyone needing scalable backend services and doesn’t want to build from scratch. There are a number of technical sessions available from I/O 2017 that cover many of the Firebase features in more detail.

Google also acquired Fabric from Twitter earlier this year and will likely merge those services into Firebase over the new several months.

A surprise announcement!

Much to the surprise of the developer community, it was announced that Kotlin will receive first-class support on Android. Kotlin is a programming language for the JVM developed by JetBrains which has been gaining in popularity with the Android community since it’s announcement in 2012. Kotlin provides a fresh alternative to Java in the Android ecosystem. Brandon Wever has discussed the Kotlin announcement in detail.

Google I/O 2017 was chock-full of exciting announcements. Google’s shift to an AI-first approach will be exciting to follow as they push the AI and machine learning fields forward.