The era begins in which AI no longer forgets who we are after closing the browser tab or the app.
Until today, moving from one digital assistant to another or simply starting a new session meant resigning oneself to a kind of tabula rasa, where every preference, interest or personal detail had to be communicated again from scratch.
Google has decided to break down this barrier by introducing in Italy a package of features for Gemini that focus entirely on continuity and on extreme personalization of the user experience.
Gemini is now an assistant that learns to know you
The centerpiece of this update is the function named Memory. Thanks to this integration, Gemini does not merely process a single ad-hoc request, but retains a record of information shared during previous interactions.
This is an evolution that turns the algorithm into a collaborator capable of developing a deep knowledge of the user’s habits. If in the past there was talk of a passion for Japanese culture or of a specific list of recommended readings on social platforms, Gemini will use these data to tailor every future response.
The activation of this system happens automatically. Users who prefer to maintain a more private approach must intervene manually in the app’s settings, navigating the dedicated Personal context menu.
Google nevertheless offers granular management tools, allowing to view, modify or delete individual memories or entire conversations via the account activity section. It is a delicate balance between the efficiency of a tailored service and the handling of sensitive data that reside on Mountain View’s servers.
The end of barriers between platforms
In addition to the internal memory, the most notable novelty concerns the simplification of the transition from other competing services such as ChatGPT or Claude.
Google has understood that the main brake on switching assistants is the loss of information accumulated over time. To address this, it has introduced two migration paths extremely intuitive.
The first consists of a prompt-based transfer system: the user receives a textual command to paste into the old chatbot, which generates an organic summary of all known preferences. Once this text is pasted into Gemini, the AI analyzes it and instantly integrates it into its own context database.
For those who want an even deeper transition, it is now possible to directly upload a ZIP archive containing the entire chat history exported by other providers. This not only keeps the machine’s "memories" alive but makes the entire history searchable and usable to resume old logical threads.
With this strategy, the company shifts the competition from pure compute power to practical usability, facilitating the entry of new users who no longer have to fear losing months of training and customization of their virtual assistant.
In an increasingly crowded market, the ability to embrace the digital legacy of competitors could prove to be the decisive move to consolidate the user base in our country.
