Google’s Gemini chatbot now features a “memory” function, designed to remember user preferences, life details, and work-related information. While this capability is currently available to a limited group of users, it has the potential to enhance conversation context. For instance, users can inform the chatbot of their favorite food, enabling it to consider gastronomic preferences when suggesting nearby restaurants.
Similar to ChatGPT’s memory feature, Gemini’s “memory” allows for more personalized and relevant interactions. However, this feature is only accessible to Google One AI Premium subscribers, costing $20 per month. At present, the feature is limited to the web version and isn’t yet available on iOS or Android apps, adds NIXsolutions.
Key Features and Privacy Assurances
Examples of supported requests include adapting language style, focusing on specific programming languages, or factoring in daily expenses for trip planning. Currently, the memory feature only supports English and can be disabled anytime. Stored memories remain until manually deleted, with Google emphasizing they are not used for model training or shared externally.
A Google representative assured, “Your stored information is not transmitted anywhere and is not used to train the model.” Despite this reassurance, computer security experts caution that memory functions in chatbots may be vulnerable to misuse if not adequately protected. For example, prior incidents with ChatGPT revealed how false memories could be introduced to exploit user data.
While the Gemini memory feature adds convenience and personalization, concerns about security remain. As the feature expands and improves, we’ll keep you updated.