Gemini Gets Personal Intelligence, Can Connect to Google Apps to Provide Personalised Answers

Google introduced a new feature in Gemini, its artificial intelligence (AI) chatbot, that solves two key problems. Memory and context are crucial for anyone who invests in a conversation with the AI or has a long-term project going on. Gemini, like any other chatbot, suffers from hallucination and forgetfulness in a chat window that goes on for too long. While users can tweak its personality slightly, it is also not consistent across the platform. But now, with Personal intelligence, the Mountain View-based tech giant is offering a unique way for users to get tailored responses from Gemini.

Google Introduces Personal Intelligence in Gemini

In a blog post, the tech giant detailed the new feature, and how it can help users. If you have ever asked an AI chatbot “Do you know my favourite colour?” and were frustrated with a random, and mostly incorrect response, this is exactly the problem Google wants to solve with Personal intelligence.

For any chatbot to become unique to a user, it needs to remember things about them and it should have the right context and information. So far, AI is not great at memory despite breakthroughs in techniques. However, it is getting increasingly better at fetching information from external data sources. Google is tapping on this capability, also known as retrieval-augmented generation (RAG) to personalise Gemini.

The feature is not a new one, but a new implementation of an existing capability. For a while, users have the option to connect Gemini with other first-party apps, such as Google Drive, Gmail, Photos, and more. So far, this capability was used for times when user needed a particular information from their app, or if they wanted Gemini to take an action based on the app (play a YouTube video or draft an email).

With Personal intelligence, this entire experience is now seamless and intuitive. Users, once they have access to it, can go to the Gemini app and tap on “For you” option. It turns on the contextual memory and the app can now answer queries about the user or their daily routine. It is also turned off by default and users can choose to turn it off by going to the Settings and visiting Connected apps at any time.

Google also stated that Gemini will try to reference or explain the information it used from a user's connected sources, making it easier for them to verify it. In case the chatbot doesn't show it directly, users can request it to share more information about the source. “You can also easily regenerate responses without personalization for a particular chat, or use temporary chats to have a conversation without personalization,” the post added.

The tech giant claimed that Gemini with personal intelligence will not make proactive assumptions about sensitive data, such as health information, unless the user asks first. Additionally, Google also clarified that no data taken from the connected apps is used to train the model.