Google, the global tech giant, has introduced a new feature aimed at detecting live scams on Android phones worldwide. This initiative was unveiled during the company’s annual developer conference on Tuesday, where it announced several AI-related updates.
During the press release, Google disclosed that the Gemini Nano, functioning as live scam detection, will utilize AI features on Android phones. It will identify scam phone calls in real time, issuing warnings to users and aiding in the prevention of fraudulent activities.
Most of the features announced for Android phones are slated to be available later this year, with initial rollout planned for Google’s Pixel devices. Additionally, Google revealed that Gemini Nano will soon be available on-device for other functionalities in the coming months, allowing Android phones to comprehend context such as sights, sounds, and spoken language.
The multimodal capabilities of Gemini Nano will be integrated into TalkBack, specifically targeting individuals with blindness or low vision.
Furthermore, Google announced updates to its Circle to Search feature, which allows users to search for pictures, text, and other information without leaving an app. This feature is being expanded to cater to the needs of students, enabling them to receive step-by-step instructions for solving math or physics problems directly from an Excel sheet or other documents.
According to Google, Circle to Search will soon have the capability to assist with more complex problems involving symbolic formulas, diagrams, graphs, and more, enhancing learning experiences for students.
At the developer conference, Google also announced upgrades to its Gemini Pro 1.5 model, which enhances the Google flagship chatbot’s ability to analyze vast amounts of data. The company emphasized its commitment to advancing the Gemini platform, stating that the Pro model will be available in 35 languages to users subscribed to the ‘Gemini Advanced’ subscription.
Furthermore, Google stated that the improved Gemini 1.5 Pro, offering up to 1 million tokens for developers globally, marks a significant advancement in long context understanding.