From Passive Consumption to Active Knowledge Construction
Knowledge is constructed, not transmitted. Learners build understanding through active engagement.
— Piaget, Vygotsky
Move from Remember → Understand → Apply → Analyze → Evaluate → Create
Wisdom begins with recognizing what one doesn't know. The best teachers ask questions.
— Socrates, 470-399 BCE
Awareness of one's own learning process. "Thinking about thinking."
— Flavell, 1979
Explore knowledge structure
Learn by questioning
Multi-agent awe
Diagnostic reasoning
158 tracked behaviours
myKE uses AI differently: to evaluate questions, trigger awe and wonder, and build diagnostic reasoning — not just answer.
Concepts don't exist in isolation—they form interconnected webs of meaning.
| Type | Visual | Meaning |
|---|---|---|
| Concept | ⬜ Rectangle | What to understand |
| Skill | 💊 Pill | What to do |
| Prerequisite | ⬡ Hexagon | What to know first |
| Requires | Must understand A before B |
| Enables | A makes B possible |
| Related | Connected themes |
"What is the answer?"
Student asks → AI answers → Learning ends
"What is the right question?"
AI challenges → Student questions → Learning deepens
Can be answered by simple lookup
Require integration of multiple ideas
Asks WHY and HOW, not just WHAT
Connects concepts together
Explores edge cases and exceptions
Contextualizes to real problems
AI creates a realistic context requiring understanding of the concept
Learner chooses 1-3 concepts to focus on, revealing what they find challenging
AI provides examples of deep questions as scaffolding (optional starting point)
Learner asks their own question or uses/modifies a suggestion
AI assesses if question is "deep" — if not, guides toward depth
For deep questions: AI guides reasoning, doesn't just give answers
Most AI tools explain. Wonderment makes you feel the weight of what you don't yet understand.
Select or write a curiosity observation about the concept
Pick a question type: paradox, inversion, scale, analogy, reframe
Compose your own answer — genuine engagement is measured
Counter-arguments challenge your thinking
A wonderment moment that reframes the concept entirely
Doctors diagnose. Engineers troubleshoot. Detectives reconstruct. These are not recall tasks — they are reasoning under uncertainty.
Scenarios span engineering failures, biological breakdowns, and social system collapses.
Systematic vs. selective clue reading reveals reasoning approach
Direct measure of Bloom's Apply & Analyze — hardest to fake
WHY it broke (mechanism) vs. WHAT broke (symptom)
| Typical AI Use | myKE AI Use |
|---|---|
| Answers questions | Evaluates questions |
| Provides information | Generates scenarios |
| Explains concepts | Guides exploration |
| Reduces effort | Requires engagement |
| Ends curiosity | Stimulates curiosity |
In Q2L, AI's most important role is evaluating whether a learner's question is "deep" and providing constructive feedback when it isn't.
Teachers bring pedagogical judgment.
AI brings comprehensive coverage.
Together: richer curricula.
Secure, private class joining
See student engagement quality
Students can rearrange nodes
"Did they get the right answer?"
"How did they engage with the material?"
The drive to ask "why?" and "how?"
Ability to evaluate claims and sources
Awareness of one's own learning process
Confidence to form and revise opinions
In a world where AI can answer almost any question, human value shifts to:
Asking the right questions • Evaluating answers • Making connections • Creating meaning
Questions about myKE? Ideas for improvement?
We'd love to hear from you.
Thank you for your attention.
Studies have shown concerning trends in how students learn with technology:
The challenge isn't access to information—it's developing the cognitive skills to evaluate, connect, and apply that information meaningfully.
| Cognitive Level | Description | myKE Feature |
|---|---|---|
| Remember | Recall facts and basic concepts | AI definitions, node descriptions |
| Understand | Explain ideas or concepts | Edge explanations, chat drilldown |
| Apply | Use information in new situations | Q2L scenarios, assessments |
| Analyze | Draw connections among ideas | Concept maps, node expansion |
| Evaluate | Make judgements and critiques | Q2L rubric, Wonderment critics panel |
| Create | Generate new ideas from parts | Wonderment synthesis, deep Q2L questions, WTB diagnosis |
The Q2L feature specifically targets the highest levels by requiring learners to create their own questions, not just respond to pre-made ones.
"A pharmaceutical company is testing a new pain medication. They've recruited 120 patients and divided them into 4 groups: placebo, low dose, medium dose, and high dose. After 4 weeks, they measure pain reduction scores..."
"What is ANOVA?"
AI Response: "That's a good start, but it's a definitional question. Can you ask something that explores why or how ANOVA works in this scenario? For example: Why is ANOVA appropriate here instead of multiple t-tests?"
"Why would conducting multiple t-tests between each pair of groups be problematic, and how does ANOVA address this?"
AI Response: "Excellent deep question! This gets at a fundamental issue in statistics. Let's think through it: When you run multiple comparisons, what happens to your overall Type I error rate?..."
myKnowledgeExplorer/
├── client/ # React SPA (Vite + Tailwind)
│ ├── src/
│ │ ├── components/ # UI components
│ │ ├── contexts/ # Auth, Learning, Tab contexts
│ │ ├── pages/ # Route pages
│ │ └── services/ # API client
│ └── public/ # Static assets
├── server/ # Express API
│ ├── src/
│ │ ├── routes/ # API endpoints
│ │ ├── services/ # Business logic
│ │ │ └── ai/ # LLM providers
│ │ └── index.js # Entry point
│ └── data/ # Local JSON storage
└── docs/ # Documentation
/api/map/generate - Generate concept map/api/map/q2l/scenario - Generate Q2L scenario/api/map/q2l/evaluate - Evaluate question depth/api/map/assessment - Generate assessments