Google is currently testing a novel method to train its artificial intelligence algorithms using Android phones. Specifically, it leverages data collected from Gboard to tailor users’ search results.
Gboard will track which search suggestions are selected and which are not. This information is then used to personalize search results on each user’s device.
Google will compile these personalized changes to create a single new update for the app for all users. So, if you use Gboard on Android, you can expect frequent updates.
However, you don’t need to wait for a new app update to start experiencing more personalized search suggestions. Your personalized algorithm will begin rolling out as soon as data is collected.
Despite initial concerns, Google assures that this method is more private. The data used to enhance an individual’s app will never leave their phone.
This is enabled by a new AI training method called “Federated Learning.”
“Federated Learning enables mobile phones to collaboratively learn a shared prediction model while keeping all the training data on the device, decoupling the ability to do machine learning from the need to store the data in the cloud.”
Google admits it has only begun to explore the potential of Federated Learning.
“Beyond Gboard query suggestions, for instance, we aim to improve the language models that power your keyboard based on what you actually type on your phone (which can have a unique style) and photo rankings based on the types of photos people look at, share, or delete.”
The data collection and corresponding updates will not affect the performance of Gboard or the phone’s battery life. Google states that these processes will only take place when the phone is plugged in, idle, and connected to free Wi-Fi.
