In-Chrome AI

Inspiration

We were inspired by the potential of Google's Gemini Nano, a powerful language model now available directly in the browser. We envisioned a Chrome Extension that leverages this readily available AI to enhance user productivity and streamline text input across various websites. The goal was to create a practical and useful application demonstrating the power of Chrome's integrated AI capabilities.

What it does

This Google Chrome Extension provides AI-powered suggestions for text input using Gemini Nano. As a user types in any editable field (e.g., text boxes, search bars, code editors), the extension analyzes the entered text and generates relevant and contextually appropriate suggestions in real-time. This helps users write faster, more efficiently, and with improved accuracy.

How we built it

The extension is built using Deno and TypeScript, leveraging the robustness and type safety of these technologies. Deno's efficient build system and TypeScript's strong typing capabilities made development faster and more reliable. The extension interacts with Gemini Nano through the browser's built-in API to receive AI-powered suggestions. The development process included using Deno's built-in tooling for building (deno task build), formatting (deno fmt), and linting (deno lint).

Challenges we ran into

One challenge involved handling diverse input types and contexts across different websites, ensuring that suggestions remained relevant and helpful regardless of the webpage.

Accomplishments that we're proud of

We successfully built a functional Chrome Extension that seamlessly integrates with Gemini Nano. The extension provides helpful suggestions in real-time, enhancing the user experience across a broad range of websites. We're particularly proud of the clean and efficient codebase that Deno and TypeScript have achieved. The project showcases the ease and speed of developing Chrome extensions with built-in AI capabilities.

What we learned

This project taught us valuable lessons in browser extension development, efficient use of AI language models, and the importance of performance optimization. We gained experience working with Deno and TypeScript in a real-world application and learned how to integrate a language model into a user-facing application effectively. The experience highlighted the immense potential of browser-integrated AI to revolutionize user interactions with web applications.

Future work

We believe that Chrome's built-in AI capabilities represent a significant step towards more intelligent and intuitive web experiences. Future development could explore enhancing context awareness, personalizing suggestions based on user preferences and history, and integrating other AI features to create even more powerful and helpful extensions. Additionally, enhancing the relevance of suggestions by supplying additional context to the language model regarding the current webpage could further improve the user experience.