AI Business is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.

Google is Testing Training its AI Directly on Our Smartphones

by Ed Lauder
Article Image

Google is testing a new way to train its AI algorithm directly on your smartphone as opposed to feeding its servers with raw data.

Google is testing a plethora of new ways to improve its AI systems. They want to create a streamlined approach to machine learning to ensure that your phone's AI is catered specifically to your needs. People are starting to get a bit worried about the amounts of data Google has on their users.

Think about it, Google offers its services like Google Drive for free, so naturally people will use them. Yet, this in turn means that they have access to considerable amounts of data on their customers. However, this new approach Google is testing, may help to dampen these fears somewhat.

Google has announced via a blog post it's new approach to training their machine learning platform. It's called "Federated Learning", and is a completely new way to train their AI by taking advantage of your phone's own processing power, as opposed to sending vast amounts of data to their servers directly via the cloud.

“It works like this: your device downloads the current model, improves it by learning from data on your phone, and then summarizes the changes as a small focused update,” wrote Google in their blog post.

“Only this update to the model is sent to the cloud, using encrypted communication, where it is immediately averaged with other user updates to improve the shared model. All the training data remains on your device, and no individual updates are stored in the cloud,” Google explained.

According to Google, they are currently testing this new approach on their keyboard app, Gboard, only on Android smartphones. “When Gboard shows a suggested query, your phone locally stores information about the current context and whether you clicked the suggestion. Federated Learning processes that history on-device to suggest improvements to the next iteration of Gboard’s query suggestion model,” Google highlighted.

Smartphone users may be slightly concerned about battery life and data usage, which is understandable. However, Google has reassured us that "Federated Learning" will only be initiated once your phone is plugged into a power source and connected to a free wifi network, which will include your home wifi. This will ensure that you don't get a huge bill from your respective network providers at the end of the month and that your phone's battery life won't be affected.

Feature image courtesy of Maurizio Pesce

Practitioner Portal - for AI practitioners

Story

Shaping AI and analytics services

8/6/2020

AI and analytics teams must market themselves – and they have to have a clear service strategy

Story

UK's ICO publishes guidance on AI and data protection

8/3/2020

The document aims to help organizations mitigate the risks of using personal data in AI applications

Practitioner Portal

EBooks

More EBooks

Upcoming Webinars

More Webinars

Experts in AI

Partner Perspectives

content from our sponsors

Research Reports

9/30/2019
More Research Reports

Infographics

AI tops the list of most impactful emerging technologies

Infographics archive

Newsletter Sign Up


Sign Up