Google’s Gemini multi-modal models included Gemini Nano, designed for mobile devices including the Pixel 8 Pro.
A Google blog post unveiled numerous new AI-powered features for Pixel phones, enabled by Gemini Nano, a specialized variant of Google’s Gemini AI model optimized for on-device operations.
Gemini Nano is extremely lightweight yet capable, meaning it can process data locally on the device rather than sending and receiving it from the cloud. Gemini Nano is powered by the Pixel 8’s Tensor G3 chip.
It offers two new cutting-edge AI-powered features to the device. The first is the ‘Summarize’ feature within the sound recorder app, which can analyze and condense recorded conversations, harnessing Gemini’s ability to process audio to text.
The second is an intelligent reply function within the Gboard app, currently tailored for WhatsApp, offering context-sensitive response suggestions.
Other features include ‘Video Boost’, a collection of video improvement tools driven by Gemini’s video processing abilities. This includes enhancements in color grading, lighting adjustment, stabilization, and grain reduction to elevate video quality.
Notably, Video Boost incorporates Google’s Night Sight technology, enabling superior video capture in low-light conditions. However, this feature requires uploading videos to Google Photos for processing.
The Pixel 8 and Pixel 8 Pro also offer Night Sight in Timelapse mode, building upon the series’ impressive low-light performance. Additionally, the Photo Unblur feature has also been refined with AI.
Other additions include the functionality to use Pixel smartphones (Pixel 6 onwards) as webcams, an upgraded Google Password Manager, and a new repair mode that protects user data during device servicing.
Not long ago, Google announced AI image editing features, including “Best Take,” which enables users to swap people’s faces post-photo. This contributed to the ongoing debate surrounding deep fakes, with some pointing out that image manipulation is now simpler than ever.
Andrew Pearsall, a seasoned photographer, explained, “One simple manipulation, even for aesthetic reasons, can lead us down a dark path.”
However, as Professor Rafal Mantiuk from the University of Cambridge rebutted, “People don’t want to capture reality…They want to capture beautiful images.
In any case, local AI is coming to smartphones, starting with the Pixel 8 Pro. It will eventually become the norm.