Google’s Gemini AI Goes Live for All US Users
▼ Summary
– Google has expanded access to its Personal Intelligence feature, making it available to all free-tier users in the US, not just paying subscribers.
– The feature automatically personalizes Gemini’s responses by using data from connected Google apps like Gmail, Photos, and YouTube.
– Personal Intelligence is currently limited to personal Google accounts and is not available for business, enterprise, or education users.
– The feature is opt-in, and users can choose to turn it off or disconnect specific apps from it at any time.
– Google states that it does not train its AI models directly on personal data like Gmail or Photos content, but on limited information such as prompts and responses.
Google has expanded access to its Personal Intelligence feature for all users in the United States, moving it from a premium subscription benefit to a free-tier offering. This update allows individuals to integrate data from various Google applications, enabling the Gemini AI to deliver more contextually aware and personalized responses. The functionality is now accessible through AI Mode in Search, the Gemini app, and Gemini within the Chrome browser, though it remains exclusive to personal Google accounts for now.
This feature operates by drawing on information from connected services such as Gmail, YouTube, and Google Photos. By doing so, it can tailor its suggestions and answers automatically, eliminating the need for users to manually provide extensive background in their queries. For instance, with personalization active, Gemini might propose shopping ideas based on recent purchases or offer technical support guidance using known device details. Early testers have noted that while the AI excels at grasping broad interests, the precision of its nuanced recommendations can sometimes falter.
Importantly, Personal Intelligence remains an opt-in service. Users who prefer not to share data from their other apps can simply leave the feature disabled. Google emphasizes that privacy controls are in place, allowing individuals to disconnect applications at any time. The company also clarifies in its announcement that Gemini and AI Mode do not directly train on the contents of personal Gmail or Photos libraries. Instead, model training utilizes limited information, such as specific prompts entered by users and the corresponding AI-generated responses.
(Source: The Verge)
