Posted on 3 June 2025
Written by Dr Sue, founder of Prompt Talk and Prompt Academy.
Designing Human-Centred Learning in an AI-Saturated World
In an age where students can generate answers in seconds, we need to ask deeper questions:
What can’t AI do — and how do we design around that?
This post is for every educator who’s ready to go beyond “AI-proofing” and start AI-aware designing. Because when we understand what AI can’t do, we discover what students still must do.
🎯 1. AI Can’t Reflect on Personal Experience
AI can write about failure. But it’s never failed. It can mimic self-reflection, but it can’t feel regret, growth, or joy.
Lecturer tip: Ask students to reflect on:
- What they learned about themselves
- How their thinking changed during the task
- What surprised them in the process
âś… Use reflection rubrics that reward authenticity, not just structure.
🧠2. AI Has No Self-Awareness
It doesn’t know it’s thinking. Students do.
Lecturer tip: Design tasks that activate metacognition:
- “What was your thought process when approaching this?”
- “Why did you make that decision?”
- “How did you know your answer was correct?”
💬 3. AI Can’t Engage in True Ethical Reasoning
AI can list values. But it doesn’t feel ethical tension or make real-life trade-offs.
Lecturer tip: Include ethical dilemmas or AI use reflections:
- “Was using AI for this step fair?”
- “How do you ensure your work remains original?”
- “Would this be acceptable in a professional context?”
❤️ 4. AI Can’t Feel
No anxiety before presentations. No pride after feedback. No motivation beyond its prompt.
Lecturer tip: Encourage students to:
- Journal emotions during the learning process
- Write learning letters to future selves
- Respond to feedback with “what I’ll improve next time”
👣 5. AI Can’t Be Held Accountable
If an answer is wrong, AI isn’t to blame. The student is.
Lecturer tip: Shift grading focus:
- From “What’s the final answer?”
- To “How did you arrive at that answer, and how do you stand by it?”
➕ Add process-based marks — for judgment, evaluation, and integrity.
🔑 Final Word
We don’t need to fight AI. But we do need to design learning that AI can’t fake.
Because students aren’t just here to produce content. They’re here to become thinkers, doers, and ethical humans — and that part is still 100% human.