The fight between operating systems (OS) is no more restricted to good interface and minor fixes. Mobile OS providers have taken it a notch up and have started experimenting with machine learning and artificial intelligence to provide a seamless user experience.
Android has upped the game and has come up with new tricks under its sleeves with its eighth major version, Oreo, launched on August 21, worldwide. The OS is using AI and ML like never before and has done a whole load of work under the hood to make Android 8 the fastest iteration of its mobile platform ever.
Android released an alpha quality developer preview on March 21. The second developer preview, beta quality, was released on May 17, and a third developer preview was released on June 8, finalising the API. It was on July 24 that a fourth developer preview was released that included the final system behaviours and the latest bug fixes and optimisations.
This release includes features that will be crucial to bringing machine learning into focus on the platform, as well as interesting additions that make the mobile ecosystem much more efficient.
A for Android, A for Artificial Intelligence:
AI is the future and Oreo has introduced a subsystem that will help enhance neural network features in the device. Google has been an ace player in developing ML and AI technologies and with its latest OS version, it is all set to bring it to every user, making the platform stronger.
Using Google's advanced machine learning techniques, Android O will bring the ability to select texts, recognise it as an address, email ID or a contact number, and will show related options. Users can long tap on a text inside the address and the OS will automatically select the complete address.
With advanced machine learning algorithms, Oreo has got the auto fill implemented better. It not only allows faster log-ins (of course, with your permission) but will also make several other things easier. The new Oreo aims to make the suggestion prompts better, letting you type the redundant details lesser.
Password synchronisation and new device detection is also made efficient with the release of Oreo. Though, it will take time for some applications to adopt it.
Other notable Android Oreo features that include some sort of ML/AI, are the following:
- Better cache data management: Every app will have a cache data limit. When the system needs to free up memory, it will automatically delete cached data from apps that have exceeded their limit.
- Smart Text Selection: This feature will use machine learning to recognize commonly copied words (such as business names, phone numbers, URLs, etc.) and quickly copy them without having to select each letter or number of the intended selection.
- TensorFlow Lite: This subsystem will help apps perform significantly faster by way of on-device machine learning. This scaled-down version of Google's machine learning tool uses Long Short-Term Memory (LSTM) to improve the overall experience.
- Google Play Protect: This feature is aimed at developers only. To make this release even more important, Google Play Protect will ensure malicious software doesn't make its way onto devices. This new security feature starts at the Google Play Store by scanning apps the second they are submitted and continues on to the device, where apps are scanned in real-time.
- Google Play Console Dashboards: It will display performance issues (such as battery drain, crashes, interface problems, etc.) to help developers resolve problems with their apps. This will greatly benefit all Android users because app problems will be resolved faster.
- Combat Notification Spam: Android Oreo lets you choose whether or not a channel can buzz your phone or jump to the top of your notifications, or lets you block channels altogether. This should help you combat notification spam.


Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.