TouchGrass AI: Less Scrolling, More Exploring with Open-Source AI

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass.

What I Built

🌿 TouchGrass AI — Less Scrolling, More Exploring

TouchGrass AI is a web application designed to encourage people to spend less time on screens and more time exploring the natural world.

The idea is simple: instead of using technology only for entertainment, we can use AI to inspire real-world experiences.

Users can generate outdoor missions, explore nature-inspired activities, and earn points for completing challenges. Examples include observing different leaves, listening to birds, exploring trees, and taking mindful walks.

The project is designed for students, nature lovers, and anyone who wants to develop healthier screen habits while discovering the world around them.

My goal is to make technology a bridge to nature, rather than a barrier.

Demo

🔗 GitHub Repository: https://github.com/sy8737971-a11y/TouchGrass-AI

🌐 Live Demo: A deployed website link will be added when hosting is available.

Code

💻 Explore the source code: https://github.com/sy8737971-a11y/TouchGrass-AI

The project uses HTML, CSS, and JavaScript to create a responsive interface, generate outdoor missions, and track completed activities.

The repository contains the implementation and documentation for the project.

How I Built It

I built TouchGrass AI using HTML, CSS, JavaScript, and Transformers.js, with the goal of making open-source AI accessible through a simple web application.

The project integrates the open-weight Xenova/flan-t5-small model to generate nature-related activity suggestions directly in the browser.

Here's how the main technologies work together:

• HTML5: Structures the website and its mission dashboard.
• CSS3: Creates a responsive, nature-inspired interface.
• JavaScript: Handles mission generation, completion tracking, points, and user interactions.
• Transformers.js: Provides the browser-based AI inference pipeline.
• FLAN-T5 Small: Generates short outdoor activity suggestions.
• Fallback missions: Allow users to continue exploring if the AI model cannot load.

The focus was to build a lightweight application without requiring a paid, closed AI API for every mission.

The current implementation requires internet access to download the AI library and model initially. Fully offline use would require additional caching and offline testing.

Why Does Open Innovation Matter?

Open innovation matters because developers should have the freedom to experiment, learn, adapt, and build on existing AI technology.

For TouchGrass AI, open-weight AI provides an opportunity to run inference on a user's device instead of depending entirely on a closed AI service. This can reduce dependence on paid inference APIs and offers a path toward improved privacy and offline functionality.

It also makes experimentation easier. Developers can explore different models, change prompts, improve mission generation, and extend the application as their ideas grow.

Most importantly, open-source tools make it easier for students and independent developers to turn ideas into working projects without needing expensive infrastructure.

I built TouchGrass AI to explore how accessible AI can encourage people to reconnect with nature and spend more time in the real world.

My Agent Session

This section is optional. I have not linked an agent session yet.

If I record a session using DevRelay, I will add the session link here.

Prize Categories

Primary theme: Touch Grass — Open-Source AI

My project focuses on using open-weight AI to encourage outdoor exploration and screen-free activities.

Thanks for checking out TouchGrass AI! 🌿

Less scrolling. More exploring. More nature.