Teach with AI.
Make room for thinking.

What is Vibe Teaching?

Vibe Teaching is Laurence Svekis’s approach to using AI to design and guide active learning. Educators set the goals, shape the experience and judge the evidence of understanding. AI supports explanations, questioning, practice and feedback. Learners remain responsible for thinking, making decisions and demonstrating what they know.

It is for teachers, tutors, trainers, facilitators and course creators who want to move beyond generating lesson content toward helping people develop real capability.

From Vibe Learning to Vibe Teaching

I’m Laurence Svekis, an educator, developer and author. I began working in online education in 2001. After seeing the emergence of vibe coding in 2025, I began asking how conversational AI could support a more structured way to learn.

I coined the term Vibe Learning for my approach to AI-assisted learning and developed the Vibe Learning Framework. Shortly afterward, I began using Vibe Teaching for its educator-facing counterpart: a method of using AI to teach, guide practice and support understanding.

My vision is to give educators more ways to respond to learners while preserving the curiosity, effort and judgment that make learning meaningful.

One philosophy. Two perspectives.

Both share a central principle: use AI to strengthen human thinking.

The educator designs the learning experience

AI can produce material quickly. The educator decides whether that material is accurate, appropriate and useful for the learning goal.

The Vibe Teaching learning loop

Use these eight moves as a flexible lesson-design loop. Return to earlier steps when learner responses reveal a gap.

1. Orient

   Define what learners should be able to do and what evidence will show it.

2. Understand

   Connect the concept to prior knowledge using explanations, examples and questions.

3. Predict

   Ask learners to commit to an idea before revealing an answer or outcome.

4. Apply

   Give learners something meaningful to solve, build, write, explain or decide.

5. Feedback

   Identify a specific gap, offer the smallest useful hint and invite another attempt.

6. Recall

   Ask learners to reconstruct key ideas without AI or reference material.

7. Connect

   Relate the new idea to another topic, experience or situation.

8. Transfer

   Change the context and check whether learners can apply the idea independently.

A classroom example: teaching JavaScript conditionals

Goal: Learners can write and explain a condition that checks whether a person is eligible for a discount.

1. Predict: Show a short if/else example. Ask which branch runs for three different inputs.
2. Explain: Ask learners to describe the condition in their own words.
3. Apply: Have learners write their own discount rule before asking AI for help.
4. Feedback: Ask AI to suggest a hint for a fictional incorrect answer. Review the hint before sharing it.
5. Recall: Close the chat and ask learners to rebuild the condition.
6. Transfer: Change the task to free-shipping eligibility and include a boundary case.

The evidence is the learner’s code, explanation and handling of a new case—not the quality of the AI-generated answer.

Three prompts educators can adapt

Replace the bracketed details. Review AI output for accuracy, suitability and alignment before using it with learners.

Help me design a Vibe Teaching lesson on [topic] for [learner level].
The learning goal is [observable capability]. We have [time].
Include a prior-knowledge question, a prediction task, a short explanation,
independent practice, feedback, recall without AI, and a transfer task.
For each step, state what the learner does and what evidence I should observe.
Flag factual claims I should verify. Do not provide solutions in the learner handout.

Act as a guided practice partner for [topic].
Ask one question at a time. First ask me to explain my current thinking.
If I make a mistake, give the smallest useful hint and let me try again.
Do not reveal the complete solution immediately.
After a successful attempt, ask me to explain why it works and try a new case.

Create an assessment for [learning goal] at [learner level].
Include an explanation question, a misconception to diagnose,
a practical task, and a new situation requiring transfer.
Separate learner questions from teacher guidance and suggested answers.
Provide observable criteria for understanding and partial understanding.
Include a short independent check completed without AI.
Use fictional examples and no identifiable learner information.

These prompts are starting points. A chatbot may not follow every instruction; the educator remains responsible for the experience.

Practical principles for teaching with AI

• Start with capability: name what learners should be able to do.
• Protect productive struggle: allow an attempt before offering help; adjust support when difficulty becomes unproductive.
• Make thinking visible: ask for reasoning, predictions and explanations.
• Adapt with purpose: offer simpler examples or harder tasks while keeping the learning goal clear.
• Verify the material: check generated facts, examples and suggested answers.
• Keep learner information private: use fictional or de-identified examples and follow your institution’s rules.
• Check independence: include recall and application without AI.

Start with one lesson

Choose a lesson you already teach. Keep its learning goal, add one prediction question, replace one immediate answer with a hint, and finish with an independent transfer task. Compare the learner’s first and final attempts to decide what to adjust next.