Google made it apparent that the Google Pixel 8 Pro was more of an AI phone than just an ordinary Android flagship when it unveiled the device earlier this year. Google is now making a stronger push toward that AI future with two new features built on the internal Gemini Nano model.
In Android 14, Google developed a new system called Android AI Core in order to bring Gemini Nano to the Pixel 8 Pro. With native support for Tensor mobile silicon, it’s Google’s most portable and effective AI model. Sensitive information is unlikely to leave your phone when everything is done locally thanks to on-device operation.
Most notably, these AI features don’t require an internet connection to function, nor does Google’s cloud. In terms of features, Gemini Nano provides some neat tricks like sophisticated grammar-correction and proofreading tools, context-aware smart responses, and advanced text summarization.
The Recorder app for Pixel phones now has a new summarizer system, which is the first Gemini Nano-based feature on the Pixel 8 Pro. The app, which is already very good at transcription, will now provide an easily digestible bullet point summary of the audio recording. While GPT-based programs and tools like Shortwave provide their own unique approaches to summarization, they are dependent on an internet connection to function.
However, Recorder’s AI summarizer system is currently only capable of summarizing text in English. Regretfully, we are unsure if additional languages and older devices will be supported in the future. It’s interesting to note that the app can now translate audio in 28 different languages.
Next, the Gboard app has a brand-new intelligent response feature. Although more apps will soon add support for Google’s keyboard app’s predictive text response feature, the feature is currently available worldwide in WhatsApp. The key to this is conversational awareness; since all processing takes place natively, privacy is preserved and suggestions are made quickly.
It’s interesting to note that Google is not restricting the AI Core to its own chips. The company claims that it can benefit from Qualcomm, MediaTek, and Samsung’s machine learning engines that are based on silicon.
That makes perfect sense, considering that both Qualcomm and MediaTek currently provide high-end and mid-range processors capable of supporting AI models with billions of parameters. Furthermore, the architecture’s inherent flexibility allows developers to customize AI-powered features and optimize it for their own apps.
Google is enhancing the Photo Unblur feature to enhance its ability to identify pets such as dogs and cats, in addition to the AI functionalities integrated into the Gemini Nano model. The Pixel phones now have a new Clean feature that can remove smudges and gunk that can be seen in scanned documents, taking an APGE straight out of the Magic Eraser book.
A new Dual Screen Preview mode on the Google Pixel Fold allows you to use the outside screen as a viewfinder. Users are also starting to have widespread access to the Pixel phone’s high-definition webcam feature.
Google is now offering a Repair Mode for Pixels, which, like Samsung, locks all user data in a secure location while the device is being repaired. In addition, when the Google Assistant is screening calls, the Pixel 6 and all subsequent phones will now display contextual quick replies, providing users with a quick way to respond to calls via text without ever picking up the phone.