RSS Amplifier

The AI School Librarians Newsletter · Jul 27, 2026

AI Summer School 2026: Week 9

0
Sign in to vote or save

The AI School Librarians Newsletter · The AI School Librarians Newsletter

Created with ChatGPT

There is a question sitting underneath everything right now.

If AI can complete the assignment, what is the assignment actually measuring?

That question is uncomfortable.

But it is necessary.

Last week, you focused on academic integrity.

How to move from detection to designing for thinking.

This week takes the next step.

Because once you accept that shift, you have to redesign the work itself.

It is no longer enough to ask:

“Can students complete this?”

You have to ask:

“Where can I actually see their thinking?”

If you cannot answer that clearly, the assignment will not hold.

Before, you could assume a finished product reflected student thinking.

Now, you cannot.

If you do not design for visibility, you are grading assumptions.

Many assignments were built for a different environment.

An environment where:

  • information was harder to access

  • writing took longer

  • generating ideas required more effort

Now, students can produce answers quickly.

But that does not mean they understand them.

One of the biggest challenges right now is this:

You cannot always tell how students arrived at their answers.

SchoolAI is designed to address that.

Instead of just generating content, it creates structured spaces where students:

  • work through ideas step by step

  • respond to guided prompts

  • interact with AI support while learning

At the same time, you can see:

  • how students are thinking

  • where they are getting stuck

  • what support they need

The goal is not better answers.

It is visible thinking.

Take an assignment you already use.

Ask:

“Where in this task would I be able to see how a student is thinking?”

Then add one step where that thinking becomes visible.

Original assignment:
Write an essay explaining the causes of World War I

Revised assignment:

  • Explain one cause of World War I

  • Compare it to another possible cause

  • Explain why you think one was more significant

  • Describe how you made that decision

Same content.

Now you can see the thinking.

Before:
Solve the equation and show your answer

After:

  • Solve the equation

  • Explain the steps you used

  • Identify one mistake someone might make and why

Same skill.

More thinking revealed.

Instead of:

Task → Completion

Design:

Task → Observable thinking → Explanation → Then evaluate

AI can support completion.

It cannot replace thinking when it is required and visible.

If students can move through an assignment without slowing down, they are not thinking deeply.

A small amount of friction is where learning happens.

Require students to show:

  • their initial idea

  • how it changed

  • what decisions they made

  • how they justified those decisions

It is not enough for students to explain their thinking after the fact.

You need moments where you can see it as it happens.

Ask:

“What part of this assignment requires students to think, not just produce?”

If that part is missing, the assignment needs revision.

  • Ask for comparison

  • Require justification

  • Build in reflection

Example:

“Explain your answer and what alternative you considered before choosing it.”

  • Adding more steps without adding more thinking

  • Making assignments longer instead of stronger

  • Focusing on output instead of process

More work is not the goal.

Visible thinking is.

Some students will:

  • still try to move quickly through tasks

  • give surface-level explanations

  • avoid deeper thinking

That is expected.

Students will do what the assignment allows.

If it allows surface-level work, that is what you will get.

  • Work that looks strong but lacks clear reasoning

  • Students who cannot explain their own examples

  • Responses that feel disconnected from class discussion

These are not proof of AI use.

They are signs that thinking may be missing.

Sometimes.

But a stronger approach is this:

Design assignments where:

  • AI can support thinking

  • but cannot replace it

Without intentional design, some students will engage deeply.

Others will complete the task quickly and move on.

Assignments determine which one happens.

Students with stronger skills will still show their thinking.

Others may rely more heavily on AI.

If we do not design for observable process, that gap grows.

Take one assignment.

Add one requirement:

“Explain how you made your decision and what other options you considered.”

When a student gives an answer, ask:

“How did you decide that?”

That one question shifts the focus immediately.

You do not need to redesign everything.

Start with one assignment.

Focus on one place where thinking becomes visible.

Add this to any assignment:

“What decision did you make in this work, and why?”

This is especially important in research.

Students may:

  • gather information quickly

  • summarize without understanding

  • rely on AI explanations

Shift the task to require:

  • comparison of sources

  • justification of trust

  • explanation of conclusions

Ask:

“Why did you trust this source over another?”

and

“What would make this source less reliable?”

That is where research becomes thinking.

Keep it simple:

“Why did you choose that answer?”

Ask:

“What choice did you make in this assignment?”

Choice reveals thinking.

Instead of asking:

“How do I make this harder?”

Ask:

“Where do students need to make and explain a decision?”

If thinking is not visible, learning is not measurable.

After one assignment, ask yourself:

“What did I actually learn about how my students think?”

If you try this, notice:

  • where students struggle to explain decisions

  • where responses stay surface-level

  • where thinking begins to deepen

That is where instruction matters most.

Start valuing:

  • explanation

  • reasoning

  • decision-making

Not just correctness.

That shift changes how students approach the work.

Next week, we move into assessment.

You will look at how to:

  • evaluate thinking, not just output

  • adapt grading practices

  • respond to AI-supported work

This week’s paid section is focused on designing assignments that hold up in an AI environment.

I am sharing:

  • a copy-and-paste prompt pack for making thinking visible

  • a template for building observable process into assignments

  • prompts for reflection and justification

  • a student-facing checklist for stronger thinking

  • one key upgrade move to strengthen any task

If you want tools you can use immediately, that is where we go next.

Read the original on aischoollibrarian.substack.com

Comments

Nothing yet. Say the first thing.

    Sign in to join the conversation.