Google is currently testing a new way to train its artificial intelligence algorithms using Android phones. More specifically, it is using data collected from Gboard to personalize users’ search results.
When Gboard suggests searches, the app will remember which ones were selected and which ones were not. That data will then be used to personalize the app’s search results on each individual’s device.
Google will collect all of these personalized changes and aggregate them into a single new update to the app for all users. So if you use Gboard on Android, expect many updates going forward.
However, you won’t have to wait for the new app update to start experiencing more personalized search suggestions. Your personalized algorithm will begin rolling out on the app immediately after data is collected.
Although it sounds like there massive amount of personal information being collected, Google says it’s actually more private to do it this way. Data used to improve an individual’s app will never leave their phone.
This is made possible thanks to a new method of AI training which the company calls “Federated Learning.”
”Federated Learning enables mobile phones to collaboratively learn a shared prediction model while keeping all the training data on device, decoupling the ability to do machine learning from the need to store the data in the cloud.”
Google says it has only “scratched the surface” of what is possible with Federated Learning.
”Beyond Gboard query suggestions, for example, we hope to improve the language models that power your keyboard based on what you actually type on your phone (which can have a style all its own) and photo rankings based on what kinds of photos people look at, share, or delete.”
The process of collecting the data and issuing changes will not affect performance of the Gboard app or the phone’s battery life. Google says these things will only take place when the phone is plugged in, not being used, and on a free wireless connection.
