AI can make students faster.
But faster at what?
The conversation around AI in education is often framed as:
Should students use AI?
I think there is a better question:
When should AI enter the thinking process?
That distinction matters.
A student receives a difficult task.
They are uncertain. They do not know where to begin.
So they ask AI.
Within seconds, the problem is explained, the structure is clearer, and a possible answer appears.
The task gets easier.
But something may also disappear:
the thinking the task was designed to develop.
The answer improved.
Did the learner?
Some friction builds capability
AI is brilliant at removing friction.
That is often useful.
But not all friction is waste.
Some of it is where learning happens.
The uncertainty before a solution.
- The failed first attempt.
- The sketch that does not work.
- The explanation that exposes a gap in understanding.
- The moment when the learner has to decide what to try next.
Remove all of that too early and we risk improving task completion while weakening independent reasoning.
We become more efficient.
But possibly less capable.
Earn the AI
So perhaps the principle should not be:
Use AI.
Or:
Do not use AI.
Instead:
Earn the AI.
Before AI contributes its thinking, the learner should contribute theirs.
- A hypothesis.
- A sketch.
- A paragraph.
- A solution.
- A first attempt.
Something.
This creates a very different relationship with AI.
Instead of asking:
Give me the answer.
The student can ask:
Challenge my answer.
That difference is enormous.
The START Framework
A simple sequence might look like this:
S — Struggle
Stay with the problem long enough to understand what is difficult.
T — Try
Produce an independent first attempt.
A — Ask
Now bring in AI to critique, challenge or extend your thinking.
R — Review
Compare the AI response with your own. What did you miss? What did AI miss?
T — Take Ownership
Decide what survives. Rewrite it. Explain it. Defend it.
The sequence matters:
Struggle → Try → Ask → Review → Take Ownership
AI is still part of the learning.
It simply enters later.
Why the first attempt matters
The first attempt gives students something AI cannot give them:
a reference point for their own thinking.
Without it, they only see the AI response.
With it, they can compare:
- This is how I thought.
- This is how AI approached it.
That gap becomes the learning.
- Perhaps the student lacked evidence.
- Perhaps their reasoning was weak.
- Perhaps the AI answer was polished but generic.
- Perhaps the student's original idea was actually better.
Now AI is not simply generating work.
It is making thinking visible.
A simple classroom rule
For selected tasks:
No AI until there is evidence of an independent attempt.
Not because struggle is automatically good.
But because some capability can only develop when the learner has something to push against.
And this may be one of the most important forms of AI literacy we teach:
Not only knowing how to use AI.
But knowing when not to use it yet.
The goal is not independence from AI.
Nor dependence on AI.
The goal is agency.
Knowing when to think alone.
Knowing when to ask for help.
Knowing what to accept.
Knowing what to reject.
And knowing when the final judgement still belongs to you.
If AI gives us the answer before we have properly met the problem ourselves, we should keep asking:
What exactly did we learn?
------------------------------------------------------------------------------------------
Like my contents? Support me, support my work? Buy me a Coffee?
------------------------------------------------------------------------------------------
#ArtificialIntelligence #AI #Education #AIinEducation #EdTech #TeachingAndLearning #FutureOfEducation #LearningDesign #CriticalThinking #StudentAgency #Teachers #GenerativeAI
No comments:
Post a Comment