Grades 6-9 Middle School Afterschool Mini AI Summer Camp

Afterschool center

Diamond Bar, Southern California, USA (suburban)

Undergraduate and graduate CS students; technical educators

• Afterschool center

Mini AI Summer Camp

5 days

6 students

Individual device

• No individual logins allowed
• No personal data collection
• Time-limited setup and teardown
• Variable Wi-Fi quality

Grades 6–9 (approx. ages 11–14)

Limited or no prior formal AI education

Some prior exposure to basic coding concepts assumed

• Develop conceptual understanding of artificial intelligence systems
• Learn how AI models learn from data and generate outputs
• Explore real-world applications of generative AI
• Build computational thinking through simulation-based coding

• Understand bias and perspective in AI systems
• Compare generative approaches (diffusion vs GAN, simplified)
• Build confidence in presenting AI-powered applications

• Not a rigorous machine learning theory course
• Not focused on mathematical foundations
• Not designed for formal assessment or certification

• AI concept instructional slides
• CodeCombat AI HackStack simulation labs
• Teachable Machine (browser-based)

• AI slides: What is AI; How AI learns & creates; Generative tools and applications
• HackStack labs: Pandemic simulation; Cupcake order form app; Weather app
• Teachable Machine: thumbs-up / thumbs-down training for simple ML models

• Conceptual AI model
• Generative system example
• Simulation-based decision system

English

• No personal data collected
• No student accounts created

5-day mini AI summer camp combining conceptual instruction with simulation-based coding labs

• Introduction to AI concepts and generative systems
• Guided exploration of CodeCombat AI HackStack labs
• Teachable Machine model training activities
• Discussion of bias, perspective, and limitations
• Creative project exploration and customization
• Informal project sharing and demos

• Human: defining goals, interpreting outputs, ethical reflection
• AI: generating behaviors, simulating outcomes, responding to trained models

• Live demos and walkthroughs
• Simplified analogies for GAN vs diffusion
• Peer discussion and collaborative troubleshooting
• Educator-guided reflection

• Abstract concepts required repeated analogies
• Debugging logic in simulation labs
• Balancing creativity with technical constraints

• Used simplified visual metaphors for generative models
• Allowed playful customization of projects
• Reduced emphasis on correctness; increased exploration

High engagement throughout; frequent laughter and spontaneous discussion

• Students explained how AI models learn from feedback
• Students articulated bias and perspective concerns
• Students compared different generative approaches at a conceptual level
• Students built creative projects (e.g., a Peppa Pig–themed Google browser clone)

• No personal data collected
• No student names recorded
• No online accounts created

• Practitioner observation
• Activity documentation
• Student project demonstrations

Middle school AI literacy; Simulation-based AI learning; Bias and ethics education

Learning sciences; Educational technology; AI literacy; Ethics in AI

• Completed

Middle school

Afterschool center

Generative AI, Simulation-based AI

Project-based learning, Exploratory learning

Low

None