Gemini can now make some use of being able to access your apps, thanks to a new feature that’s still in the works.
Google has announced a new feature for its AI assistant, Gemini, which allows it to provide users with even more personalized answers.
The company is releasing Personal Intelligence, a beta feature available to users in the US.
Users can now link apps such as Google Photos, Gmail, YouTube, and Search for it to provide them with answers tailored to their personal needs.
Even though Gemini can already access information from these apps, it can now utilize user data to send more tailored results.
For example, with the information found in Photos, the AI assistant helps retrieve one’s license plate without them having to go outside, or informs the user about the type of tires they need, based on the places they visited and captured in their pictures.
To make this possible, the information is “securely” taken from apps like Google Photos and Gmail.
According to the company, the feature is turned off by default, allowing users to decide when to enable it, as well as which apps to link to the AI.
When it comes to user privacy, Google says that “because this data already lives at Google securely, you don't have to send sensitive data elsewhere to start personalizing your experience.”
Gemini will also “try” to explain what kind of information it used to form answers. If it doesn’t provide references, users can ask it for more information or correct the assistant.
Google also notes that “Gemini aims to avoid making proactive assumptions about sensitive data,” such as one’s health, unless the user explicitly wants to discuss it.
Considering the considerable debate surrounding AI training, Google states that it doesn’t use users’ personal data to train Gemini.
“Gemini doesn’t train directly on your Gmail inbox or Google Photos library,” notes Google, adding that it uses specific prompts and Gemini responses for training to improve its performance.
Have thoughts about this topic? Others do, too. Join them in the discussion.
The company also warns users that, still in beta, the feature may provide them with inaccurate responses or “over-personalization,” so users are invited to provide feedback about the feature.
The feature is available for Google AI Pro and AI Ultra users in the US. It is accessible on Web, Android, and iOS.
The company states that the feature will also be available in the free tier in the future, with more countries to follow.
Unlock more exclusive Cybernews content on YouTube.
Your email address will not be published. Required fields are markedmarked