California-based Google has rolled out a new update to its Gemini that can now offer more personalised images. This feature is powered by the AI platform’s popular Nano Banana AI model, which will help users to generate visuals based on their choices and preferences without typing much detailed and repetitive prompt. In simple words, you will not be required to explain your suggestions multiple times to Gemini. Now it can understand your interests and work accordingly. For example, a basic prompt to ‘design my home’ may now involve elements like a swimming pool or a gym, based on your preferences.
How Gemini’s Nano Banana Will Use Other Google Apps For Context
This feature will work by analysing data from connected Google apps like Google Photos and Gmail. This will help the AI understand you better personally, including your activities, interests, and even labels like ‘Family.’ For instance, you can ask Gemini to generate an image of your family doing your favorite activity. It may use the available context based on your recent email with your family, supposedly about an outing, and could produce the results based on the same.
Transparency and Privacy Issues
Google has highlighted that users will be able to check how Gemini is building these personalised images. A “sources” button will appear to see what information has been used to create the visuals. If the context is wrong, you can provide feedback to the AI for enhanced results. Moreover, Google claims that there are no changes in the privacy policies. It suggests that it is not training its AI models based on user photos from Google Photos. However, the choice to connect your Gmail and Google Photos to your Gemini app is totally optional, and users can even choose to turn off the personalised images anytime.
Related News |
Gemini App Gets Personal Intelligence In India: Is It Free And Who Can Use It?
Who Can Create Personalised Images in Gemini?
This feature is currently being rolled out to select Google AI subscribers, including the Ultra, Pro, and Plus users in the US. Google will expand it to more users and regions in the future.
