Urgent.News

What's breaking now, across thousands of outlets.

AI

Touch Grass Challenge — AI Gives You a Reason to Go Outside

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass What I Built outside | AI plans. You go. It is a local-first outdoor challenge planner powered by an open-weight AI model. You tell it: where you are what you enjoy how often you want to go outside how much time you have each day what you want to get out of the experience The AI then creates a personalized set…

The "Touch Grass Challenge" is a submission for the Hacktoberfest Open-Source AI Challenge Week 1, designed to encourage users to venture outside. This local-first outdoor challenge planner utilizes an open-weight AI model to generate personalized outdoor activities based on user preferences. By simply providing information such as their location, interests, desired frequency of outdoor time, and available daily duration, the AI constructs a tailored set of challenges for the current month.

The app goes beyond simply providing a chatbot experience; it offers users engaging activities to participate in, such as taking a different route home, observing seasonal signs, exploring unknown locations, watching sunsets, or identifying unfamiliar plants. Users can complete these challenges incrementally, leaving notes about their experiences and tracking their progress throughout the month.

Once all challenges are completed, the AI generates a reflection based on the actual experiences and notes provided by the user, offering insights into their outdoor journey.

The application's architecture is built around local AI inference, ensuring independence from remote AI APIs. The technology stack includes Next.js with TypeScript and Tailwind CSS for the frontend, NestJS for the REST API, PostgreSQL and Prisma for the backend database, and Ollama with the configured model (Gemma 4) for local AI processing.

By running the AI model locally on the user's machine, the "Touch Grass Challenge" ensures that users can generate personalized outdoor plans without needing an internet connection, emphasizing the value of an offline AI experience in fostering real-world engagement.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

AI hackathon: how to test a solution before the final pitch

A team shows a successful model response. A judge changes the input text and gets a different result: a date disappears, an invented city appears or the request hangs.

  • Establish evaluation protocol with testing examples, comparison rules, and post-run data.
  • Provide 30 artificial announcements for development, 12 held out for final evaluation.

ButterflyBench: I Changed One Instruction. What Else Did the AI Change?

This is a submission for the Kaggle Benchmarking Challenge What I Benchmarked One of my test prompts said: "Read JSON files instead of CSV." In more than half of my reruns, three models answered by…

  • ButterflyBench measures AI model changes with small instructions
  • Author tests 40 scenarios on 20-setting command-line tool
  • Some models struggle with undoing earliest changes

More from Wednesday 7 October →