Google Gemini's new features turn generic AI into a personalized assistant across apps

Google Gemini Introduces Personal Intelligence Feature

Google Gemini has launched Personal Intelligence, a feature designed to deliver personalized AI responses by connecting data from Gmail, Google Photos, Search history, and other Google services 1

1

. Available to both free and paid users in the US, the feature works exclusively with personal Google accounts and aims to eliminate the need for users to repeatedly provide context when interacting with AI 1

1

.

The Personal Intelligence feature transforms how users interact with the chatbot by pulling relevant information from connected apps. In practical testing, when prompted to suggest summer toys for a child without specifying age or gender, Gemini correctly identified the user's four-year-old daughter, their location, and suggested age-appropriate options including water tables and inflatable toys based on past purchases and preferences 1

1

. The system even recognized brand preferences and shopping habits, recommending products from frequently used retailers like Walmart and Amazon 1

1

.

In another test through Google Chrome's Ask Gemini button, the AI identified a user's 2017 Ram 1500 Quad Cab without explicit mention and provided accurate tire size information, likely drawing from Google Photos images, Gmail messages, or past Gemini chats 1

1

. This level of personalization represents a shift from generic AI outputs to responses tailored to individual user data and preferences.

Gemini Sidebar in Google Docs Streamlines Workflow Efficiency

The Gemini sidebar in Google Docs has emerged as a practical tool for reducing workflow interruptions 2

2

. Users can now summarize lengthy documents, generate images, and receive editing feedback without opening additional tabs or leaving their workspace 2

2

.

When asked to summarize a technical document about Intel processors, the sidebar instantly broke down key points including hybrid architecture details, P-cores and E-cores, and pros such as single-core performance alongside cons like extreme heat production 2

2

. The AI writing editor also caught style inconsistencies, flagging sections where formal and casual writing mixed inappropriately and identifying shifts between first and second-person perspectives 2

2

.

The sidebar's image generation capability allows users to create custom visuals through natural language queries. One user requested a high-end PC build centered around an Intel Core i9, receiving multiple options to insert directly into their document 2

2

. However, users acknowledge that AI handling desktop-level task management requires verification, as Gemini can still make mistakes that become easier to miss when fatigue sets in 2

2

.

Gemini Integration in Google Maps Enables Contextual Search

Gemini integration in Google Maps introduces conversational search through the "Ask Maps" interface, allowing users to describe what they're looking for using natural language queries rather than keyword searches 4

4

. Instead of searching for "cafés" and manually filtering results, users can now request "a quiet café with Wi-Fi and charging points for working" and receive contextually relevant recommendations 4

4

.

The system interprets intent by analyzing reviews and user mentions of specific attributes like noise levels or crowding 4

4

. During navigation, users can interact with Gemini without disrupting their route, asking for coffee stops while the app factors in current traffic, time constraints, and location to provide immediate suggestions 4

4

.

Real-world testing demonstrates the feature's utility for discovering local hidden gems. Using prompts like "Show me under-the-radar spots in [neighborhood] that locals love but tourists usually miss" or "Find me a record store nearby that specializes in rare vinyl," users have uncovered highly-rated establishments with fewer than 200 reviews 5

5

. The voice-activated interface works seamlessly during walking routes, turning navigation into an interactive exploration tool 5

5

.

Agentic AI Capabilities Signal Shift Toward Autonomous Task Management

Leaked testing reveals Google is developing agentic AI features for Gemini Enterprise that move beyond single-task interactions 3

3

. TestingCatalog spotted a new "Agent" tab featuring a Tasks dashboard with dedicated sections for Goals, Agents, and Files, indicating Google's intention to handle multi-step tasks autonomously 3

3

.

This workspace design allows users to set broad goals while the agent breaks down tasks and uses connected services to complete them without constant input 3

3

. The interface includes panels for Connected apps and Files, alongside a "Require a human review" toggle that pauses execution before sensitive actions, suggesting Google is preparing for serious desktop-level task management where errors carry consequences 3

3

.

Early previews indicate support for reusable Skills and repeating schedules, enabling ongoing workflows that run in background while users focus elsewhere 3

3

. This positions Google in direct competition with Claude Cowork and signals the next phase of AI evolution from reactive chatbot to proactive coworker capable of managing complex workflows across multiple applications.