All big smartphone operating systems now comes with a built-in AI system. These digital assistants are designed to help the users make quick and fast decisions. Every AI platform uses the machine learning technology. The companies required to train their AI systems to work flawlessly for every individual user.
The problem with this process is it is very centralized. What that mean is your smartphones will collect your personal data and transfer it to the dedicated centralized servers. Companies like Google and Apple then use this data to train their digital assistants.
Google is planning to solve this problem by using a new type of machine learning algorithm that won’t utilize using the centralize servers. Google is calling it Federated Learning. The idea is to train your digital assistant locally using your smartphone’s processing power. It will be less time consuming and will also protect user privacy.
“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. This goes beyond the use of local models that make predictions on mobile devices (like the Mobile Vision API and On-Device Smart Reply) by bringing model training to the device as well.”
Google explained this technique saying, “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. 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 is currently testing Federated Learning in Gboard on Android, the Google Keyboard. 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.
Last year Apple also pitched the same approach called “differential privacy” to achieve a similar goal.
Stay tuned for the updates.
Recommended by: Zergnet