AI Session - National Louis University01 / 22

AI Session

Pedagogical Friction and AI

How generative AI can create unproductive success, why tertiary algorithmicity changes the work of learning, and how educators can preserve the human effort needed for deep understanding.

Opening case02 / 22

Two Students, Same Product, Different Learning

Student A

Uses AI to enter the task

Clarifies the prompt, asks for counterexamples, revises an argument, and can explain the choices.

Student B

Uses AI to exit the task

Generates a polished submission, changes a few words, and cannot explain the reasoning behind it.

Reframing the debate03 / 22

The Current AI Debate Is Too Small

Ban or embrace?

A policy question.

Cheating or innovation?

An integrity question.

What happens to learning?

The developmental question.

The real question04 / 22

What Happens When AI Can Produce the Work Students Used to Do to Learn?

The educational issue is not whether AI can make schoolwork easier. It is whether ease removes the cognitive work the assignment was designed to develop.

Session map05 / 22

Three Session Questions

1What changed? How does AI reshape the symbolic environment students learn inside?

2What is at stake? When does AI support development, and when does it bypass it?

3What should educators design? How do we reduce barriers while protecting thinking?

Ong in plain language06 / 22

Why Media Matter: Ong in Plain Language

Speech

Knowledge lives in memory, rhythm, repetition, and social presence.

to

Writing and print

Knowledge can be stored, inspected, revised, abstracted, and argued over.

Media are not just channels. They change the habits of mind a culture practices.

The analytical hinge07 / 22

Three Assumptions That No Longer Hold

Human origin

Symbolic content no longer has to begin in human consciousness.

Visible distribution

What reaches attention is sorted by opaque, personalized systems.

External media

The environment now reflects the learner's patterns back to them.

First extension08 / 22

Algorithmic Secondary Orality

Human-created, algorithmically distributed.

human createsfeed sortsattention changes

After the feed replaces the shared schedule, two people open the same platform and enter different symbolic worlds.

Second extension09 / 22

Tertiary algorithmicity

A media environment in which algorithmic systems both curate and generate symbolic content, making human authorship optional at scale.

Three effects10 / 22

Three Effects: Bypass, Simulation, Abstraction

Bypass

Thinking work is performed outside the learner.

Simulation

Dialogue feels responsive without genuine contestation.

Abstraction

Text can express positions without lived stakes.

Educational stakes11 / 22

The Educational Stakes: Not Just Integrity

Academic integrity asks, "Who produced this?" Learning design asks, "What thinking did the student have to do?"

Kapur's learning science lens12 / 22

Unproductive success

Correct-looking academic performance without the cognitive struggle required for durable understanding.

Unproductive success

Correct output, weak understanding.

Productive success

Correct output, understanding aligned.

Unproductive failure

Wrong output, little growth.

Productive failure

Struggle that prepares future learning.

Learning science foundation13 / 22

Why Struggle Matters for Learning

Retrieve: bring knowledge back without simply looking it up.

Generate: produce an idea before seeing the answer.

Revise: compare, explain, and improve thinking over time.

Response framework14 / 22

Pedagogical friction

The intentional design of useful difficulty so students still do the thinking, explaining, revising, and meaning-making that learning requires.

The framework15 / 22

Four Forms of Friction: Head, Room, World, System

Head

Noetic friction
Thinking work.

Room

Rhetorical friction
Dialogue work.

World

Existential friction
Ownership, presence, and embodied accountability.

System

Infrastructural friction
Systems that protect learning.

The equity test16 / 22

Productive vs. Exclusionary Friction

Productive friction

Builds capacity
Prepares future learning
Protects thinking

Exclusionary friction

Blocks access
Measures the wrong thing
Preserves inequity

The same AI tool can reduce both. That is why professional judgment matters.

Decision tool17 / 22

AI Use Decision Tool: Reduce Barriers, Protect Thinking

Use AI to reduce barriers when...

The barrier is not the learning objective.
The student remains responsible for meaning.
The tool improves access to the task.

Limit AI when...

The tool performs the target thinking.
The process becomes invisible.
The student cannot explain or defend the work.

Start with the learning objective. Then decide whether AI removes an access barrier or bypasses the thinking the assignment was meant to develop.

Activity18 / 22

12

Activity: Redesign One Assignment

1Identify the thinking the task is supposed to develop.

2Mark where AI could bypass, support, or deepen that thinking.

3Revise the task to reduce barriers while protecting learning.

Implications19 / 22

Curriculum, Advocacy, Policy

Curriculum

Make thinking visible before, during, and after AI use.

Advocacy

Name productive friction as a protected educational value.

Policy

Close the learning-science gap in AI guidance.

Research trajectory20 / 22

Toward the Dissertation

How do educators navigate the preservation of pedagogical friction under conditions of tertiary algorithmicity, and where does the productive/exclusionary boundary fall for different learners?

Mixed-methods case study: teacher practice, institutional conditions, student experience, and the equity boundary around AI-supported learning.

Closing claim21 / 22

Closing Claim

AI should reduce barriers to learning, not replace the thinking learning requires.

Discussion22 / 22

Discussion

1Where have you seen AI deepen learning rather than bypass it?

2Which forms of friction are hardest to protect in your context?

3What would policy look like if learning, not only integrity, were the center?