Vibe Teaching — Issue #5
Designing AI-Resilient Assignments
Let’s be honest.
Most traditional assignments are now AI-solvable in seconds.
Essays.
Summaries.
Basic coding tasks.
Definitions.
Explanations.
Students don’t even need to think anymore.
And that creates a problem.
Because if AI can complete the assignment…
What exactly is being assessed?
The Wrong Goal: “AI-Proof” Assignments
Some educators are trying to design AI-proof assignments.
Assignments where:
• AI can’t be used
• AI is restricted
• AI is banned
But this approach has two major issues.
1. It’s unrealistic
AI is everywhere. Students will find ways to use it.
2. It misses the opportunity
AI is not just a challenge.
It’s a powerful learning tool.
Instead of trying to eliminate AI…
We should design assignments that work with it.
The Right Goal: AI-Resilient Assignments
AI-resilient assignments don’t avoid AI.
They leverage it.
They are designed so that even if students use AI:
• thinking is still required
• reasoning is still visible
• understanding is still demonstrated
This is the core shift in Vibe Teaching.
The Key Question
Before designing any assignment, ask:
If a student used AI to complete this…
Would they still need to understand the concept?
If the answer is no…
The assignment needs redesign.
The 4-Part AI-Resilient Framework
Here’s a simple structure you can apply to almost any assignment.
1️⃣ Generate
Students use AI to produce an initial answer.
Example:
“Use AI to generate an explanation of how inflation works.”
2️⃣ Critique
Students analyze the AI output.
Ask:
• What is missing?
• What is oversimplified?
• What is unclear?
• What assumptions are made?
3️⃣ Improve
Students refine the response.
They:
• rewrite sections
• add examples
• incorporate sources
• clarify explanations
4️⃣ Reflect
Students explain their thinking.
They describe:
• what they changed
• why they changed it
• what they learned
This structure ensures:
AI can assist.
But learning cannot be skipped.
Example: Before vs After
❌ Traditional Assignment
“Explain the causes of the French Revolution.”
AI can complete this instantly.
✅ AI-Resilient Version
Step 1
Use AI to generate an explanation of the French Revolution.
Step 2
Identify three weaknesses or missing perspectives.
Step 3
Add one source that challenges or expands the explanation.
Step 4
Rewrite one section to improve accuracy or depth.
Step 5
Write a reflection explaining what the AI got wrong.
Now students must:
• analyze
• question
• investigate
• improve
• explain reasoning
This is learning.
The “Process Over Product” Shift
Traditional assignments focus on:
Final answers.
AI-resilient assignments focus on:
Thinking process.
You’re no longer just evaluating what students produce.
You’re evaluating:
• how they think
• how they question
• how they improve
• how they justify decisions
This is much harder to outsource.
Make Thinking Visible
One of the biggest improvements you can make is requiring students to show their thinking.
Ways to do this:
• ask for reasoning steps
• require reflection summaries
• include AI interaction logs
• have students explain decisions
• use revision comparisons
When thinking is visible, learning is measurable.
Practical AI Prompts for Teachers
Here are prompts you can use to design stronger assignments.
Prompt — Upgrade an Assignment
“Redesign this assignment so it requires critical thinking even if students use AI: [paste assignment]”
Prompt — Add Reflection
“Create reflection questions that require students to explain how they improved an AI-generated response.”
Prompt — Generate Critique Tasks
“Create a task where students must evaluate and improve an AI-generated answer on [topic].”
Prompt — Add Depth
“Suggest ways to make this assignment require deeper reasoning and analysis.”
A Simple Rule
If AI can complete the assignment in one step…
Your assignment needs more steps.
Exercise for Educators
Take one assignment you currently use.
Apply this transformation:
Old:
“Answer the question.”
New:
“Generate → Critique → Improve → Reflect”
This one change can transform how students engage with learning.
The Opportunity
AI is forcing a long-overdue shift in education.
From:
• memorization
• repetition
• output
To:
• reasoning
• evaluation
• understanding
Assignments are no longer about proving students can produce answers.
They are about proving students can think about answers.
Final Reflection
AI doesn’t make assignments obsolete.
It reveals which assignments were never measuring learning in the first place.
The future of education belongs to educators who design work that:
requires thinking, even in an AI-powered world.
That is what AI-resilient teaching looks like.
Coming Next Issue
Teaching Students How to Use AI (Properly)
Because if students don’t know how to prompt…
They won’t learn effectively.
We’ll explore:
• the CLEAR prompting framework
• common mistakes students make
• how to teach AI literacy in any subject