Hands-On AI: Education Guide¶
"LLMs made simple for students and educators"
This guide provides educational rationale, classroom implementation strategies, and learning objectives for using Hands-On AI in various educational settings.
Educational Value¶
ChatCraft was designed with specific educational benefits in mind:
- Removing Technical Barriers - Students can experiment with AI without getting lost in API keys, tokens, or complex setup
- Focusing on Core Programming Concepts - Function calls, variables, conditionals, and user input are reinforced
- Encouraging Creative Expression - Personality bots invite students to think about tone, voice, and creative writing
- Making AI Approachable - Simplifies complex technology into manageable, fun interactions
- Building Critical AI Literacy - Students learn about AI capabilities and limitations through hands-on experimentation
Curriculum Integration¶
ChatCraft can be integrated into various levels of computer science education:
Beginner Level (No Prior Programming Experience)¶
- Objectives: Basic syntax, function calls, variables
- Activities:
- Modify existing bot personalities
- Create simple conversation scripts
- Experiment with different prompts
Intermediate Level (Some Programming Experience)¶
- Objectives: Conditionals, loops, functions, error handling
- Activities:
- Create custom bot personalities from scratch
- Build a menu-based chatbot interface
- Develop a simple Q&A system
Advanced Level (Proficient in Python)¶
- Objectives: API integration, state management, more complex applications
- Activities:
- Create multi-bot conversation simulations
- Build a chatbot with memory/context
- Extend ChatCraft with additional backends
Classroom Implementation Models¶
1. Guided Tour (1 Hour)¶
A quick introduction to AI concepts through ChatCraft: - 10 min: Introduction to LLMs and ChatCraft - 15 min: Demonstration of pre-built bots - 25 min: Hands-on with simple modifications - 10 min: Discussion and reflection
2. Mini-Project Workshop (2-3 Hours)¶
Students build a specific application with ChatCraft: - 15 min: Introduction and demo - 30 min: Planning and design - 90 min: Development time - 15 min: Showcase and sharing
3. Multi-Week Unit (2-3 Weeks)¶
Deeper exploration of AI concepts and applications: - Week 1: Introduction to ChatCraft and basic interactions - Week 2: Building custom personalities and applications - Week 3: Culminating projects and presentations
Learning Objectives by Domain¶
Computer Science¶
- Understand function calls and parameters
- Practice string manipulation and text processing
- Implement basic control structures
- Gain exposure to API concepts
AI Literacy¶
- Recognise capabilities and limitations of LLMs
- Understand the role of system prompts in shaping responses
- Identify patterns in AI-generated content
- Develop critical thinking about AI outputs
Creativity & Communication¶
- Design unique bot personalities
- Craft effective prompts for desired outcomes
- Create narrative-based interactions
- Express ideas through multiple communication styles
Assessment Ideas¶
- Portfolio: Collection of bot personalities created by students
- Project: Functional application built with ChatCraft (quiz bot, story bot, etc.)
- Reflection: Written analysis of AI strengths/limitations observed
- Presentation: Demo of custom bot and explanation of design choices
Classroom Setup Requirements¶
Hardware¶
- Standard classroom computers or laptops
- Internet connection (if using remote LLM APIs)
- Or: Sufficient local processing power for running Ollama locally
Software¶
- Python 3.6+
- ChatCraft library installed
- Ollama or alternative LLM backend
Preparation Steps¶
- Install Python on all classroom machines
- Install ChatCraft (
pip install chatcraft
) - Set up Ollama with desired models (llama3 recommended for beginners)
- Test a basic example before class
Ethical Considerations¶
When introducing AI in the classroom, consider discussing:
- Attribution: When is AI-generated content appropriate to use?
- Transparency: The importance of disclosing when AI has been used
- Critical evaluation: Not accepting AI outputs without verification
- Appropriate uses: Setting boundaries for how the tool should be used
Additional Resources¶
- mini-projects.md - Ready-to-use classroom activities
- project_browser.html - Interactive offline browser for mini-projects
- bot-gallery.md - Library of pre-built personality bots
- classroom-setup.md - Detailed technical setup instructions
Example Lesson Plan: "Personality Bot Workshop"¶
Duration: 90 minutes Level: Beginner to Intermediate Objective: Students will create and interact with custom personality bots
Materials¶
- Computers with Python and ChatCraft installed
- Handout with personality prompt examples
- Worksheet for planning bot personalities
Lesson Flow¶
- Introduction (15 min)
- Demonstrate existing personality bots
- Explain the concept of system prompts
-
Show how personality affects responses
-
Planning Phase (15 min)
- Students brainstorm bot personalities
- Complete worksheet defining tone, style, quirks
-
Share ideas with a partner for feedback
-
Implementation (30 min)
- Code their personality bot function
- Test with various prompts
-
Refine system prompts based on results
-
Challenge Extension (15 min)
- Create a conversation between two different bots
-
Implement a simple menu to switch between bots
-
Showcase & Reflection (15 min)
- Students demonstrate their bots
- Class votes on most creative, most helpful, most amusing bots
- Discussion: What did you learn about how system prompts affect responses?
Assessment¶
- Completed personality bot function
- Creative application of system prompts
- Thoughtful reflection on bot behaviour
Note: ChatCraft is designed to be used with age-appropriate models and settings. Always review school policies regarding AI tools before implementation.