In this Issue: Discover how BoodleBox is leveling up with access to more powerful AI models for educational bots. Learn what really happens to your deleted chats across different AI platforms and important privacy considerations. Explore how to build reliable AI knowledge stacks for consistent educational results. Plus, get the latest on Z.ai and must-read articles on AI in education.

BoodleBox Makes Powerful Models Available | What Happens to Deleted AI Chats? |
Building Your Knowledge Stack | Tech Alert | Must-Read Articles

BoodleBox Makes Powerful Models Available for Bots

🔥 The Big Idea:

BoodleBox now offers Premium users access to more powerful AI models including Gemini 2.5 Pro, Claude 4.0 Sonnet, OpenAI 4o and 4.1, and LLAMA-4-Maverick, each with specific strengths for different educational applications.

✅ Putting It into Practice:

• Use Gemini 2.5 Pro for processing large educational materials like textbooks and curriculum guides
• Leverage Claude 4.0 Sonnet for tasks requiring nuance and ethical reasoning
• Utilize OpenAI 4o for quick content generation and general questions
• Choose OpenAI 4.1 for complex tasks requiring in-depth analysis
• Employ LLAMA-4-Maverick when developing privacy-conscious educational materials

What Happens to Deleted AI Chats?

🔥 The Big Idea:

Different AI platforms handle deleted chats differently, with BoodleBox permanently removing them from your account, while ChatGPT schedules deletion within 30 days (with potential legal exceptions), and Gemini potentially retaining anonymized versions for up to three years for training purposes.

✅ Putting It into Practice:

• Use BoodleBox when complete data deletion is a priority for your educational work
• Consider organizing important conversations into folders instead of deleting them
• Turn off chat history-saving or opt out of training when using ChatGPT for sensitive discussions
• Adjust Gemini Apps Activity settings to control how long your conversations are saved
• Be aware that legal proceedings may affect data retention policies across all platforms

Building Your Knowledge Stack: Creating Reliable #AI Systems for #Education

🔥 The Big Idea:

Educators can build reliable “knowledge stacks” by creating custom instructions for AI tools, transforming generic AI into education-specific systems that consistently deliver relevant content for lesson planning, differentiation, and assessment.

✅ Putting It into Practice:

• Create custom instructions that align AI outputs with educational standards and objectives
• Organize curriculum materials into reusable knowledge bases
• Build Custom GPTs/Bots that support specific subjects and grade levels
• Develop prompt templates that generate differentiated materials for diverse learners
• Use frameworks like VISOR or CRAFT to structure your custom AI bot development
• Consider data privacy levels when choosing between AI platforms (Local AI > Cloud API > Website/App)

Tech Alert

📢 Z.ai: New AI Chat Platform

Z.ai is now available as an AI chat platform for educators and students. This tool offers a new option for those seeking AI assistance with educational tasks. Try it out at https://chat.z.ai

Must-Read Articles

A curated selection of insightful articles on AI, learning, and classroom innovation:

Genie 3: A New Frontier for World Models

The Value of AI in Today’s Classrooms

Deciding vs. Choosing: AI and Learning

ChatGPT is a Data Privacy Nightmare

Flip It: Educational Innovation

---

Another Think Coming – MGuhlin.org
MGuhlin.org