TouchGrass AI

Title: TouchGrass AI - A Local-First Outdoor Activity Companion for Reducing Screen Time

TouchGrass AI is an open-source AI application created for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass. This innovative tool transforms screen time into practical outdoor missions by offering personalized suggestions for activities such as mindful walks, nature exploration, bird listening, gardening, and gentle movement. The primary objective of TouchGrass AI is to make the screen the shortest part of the experience, encouraging users to reconnect with the world outside.

The application utilizes Node.js and Express on the backend, paired with HTML, CSS, and JavaScript for the frontend. Ollama, an open-weight model, is leveraged for local inference, enabling the generation of realistic outdoor missions based on user activity preferences. These preferences include mood, available time, energy level, and preferred activities. The application falls back to its built-in planner in case the Ollama model is unavailable, ensuring uninterrupted user experience.

TouchGrass AI places emphasis on open innovation, as it allows users to share their mood, routines, and personal preferences without compromising their data security. By employing an open-weight model with local inference, the application reduces dependence on external services, offers users more control over their data, and facilitates customization through the inspection and modification of prompts and models.

This open-source approach also keeps the core functionality accessible even if a model is not installed or is unavailable.

In essence, TouchGrass AI is not about keeping users engaged with technology for longer; rather, it aims to use technology to help people step away from it. By fostering a deeper connection with the natural world, this local-first outdoor activity companion empowers users to prioritize their well-being and experience the benefits of spending time outdoors.