AI-Enhanced Instruction

The Opportunity and the Risk

Artificial intelligence is transforming education. Large language models can generate lesson plans, create practice problems, provide feedback, differentiate instruction, and produce learning materials at unprecedented scale and speed.

This creates both opportunity and risk.

The Opportunity:

• Generate high-quality instructional materials rapidly
• Create multiple variations for differentiation
• Provide immediate, personalized feedback
• Scale individualized instruction
• Free instructor time for higher-value activities

The Risk:

• AI-generated content without pedagogical grounding
• Cognitive offloading—learners using AI to avoid thinking
• Material that presents information without ensuring learning
• Over-reliance on AI responses without verification
• Missing the human elements that engage and motivate

Gagné's framework provides the crucial bridge: pedagogical structure that ensures AI-generated content actually supports learning, not just information delivery.

---

AI as Instructional Designer's Tool

The most powerful use of AI in education is not replacing instruction but augmenting the design process. With Gagné's framework as scaffolding, AI can rapidly generate:

Materials for Each Event:

Event 1 (Attention): "Generate five attention-grabbing hooks for a lesson on [topic], suitable for [audience]. Include one provocative question, one surprising statistic, one brief scenario, one challenge, and one multimedia concept."

Event 2 (Objectives): "Write three performance-based learning objectives for [topic] using measurable verbs. Include success criteria students can use for self-assessment."

Event 3 (Recall): "Create five retrieval questions to activate prior knowledge about [prerequisite concepts] before teaching [new topic]."

Event 4 (Content): "Write a 10-minute mini-lesson script on [topic] for [audience]. Chunk into two segments with a processing activity between. Include a worked example."

Event 5 (Guidance): "Create learning scaffolds for [topic]: one graphic organizer, five sentence starters, and a mnemonic device for remembering key steps."

Event 6 (Practice): "Generate 10 practice problems on [topic] with progressive difficulty: 4 basic, 4 intermediate, 2 challenging."

Event 7 (Feedback): "For each common error students make on [topic], write specific feedback that explains why the error occurred and guides toward the correct understanding."

Event 8 (Assessment): "Create a 10-question assessment aligned to [objectives]. Include a mix of recognition and application items. Provide an answer key with explanations."

Event 9 (Transfer): "Create three transfer tasks that require applying [concept] to novel contexts different from the learning examples."

---

Worked Example: AI-Assisted Lesson Design

Topic: Teaching CER (Claim-Evidence-Reasoning) paragraph writing to Grade 9

Event 1 Prompt: "Create two contrasting paragraphs about whether social media is beneficial—one that uses evidence well and one that makes unsupported claims. Write these as materials for a 'spot the difference' opening activity."

Event 2 Prompt: "Write learning objectives for a 45-minute lesson on CER paragraphs for Grade 9. Include success criteria as a student checklist."

Event 3 Prompt: "Create five questions to help students recall what they know about using evidence in writing before introducing the CER framework."

Event 4 Prompt: "Create a 10-minute mini-lesson script teaching CER paragraph structure. Include:

• Simple definitions of claim, evidence, reasoning
• A color-coding scheme (Claim=blue, Evidence=green, Reasoning=orange)
• A worked example paragraph with color tags
• Three teacher think-aloud lines modeling reasoning"

Event 5 Prompt: "Create scaffolds for CER writing:

1. A fillable CER paragraph organizer
2. Sentence stems for claim, evidence integration, and reasoning (at least 4 each)
3. Two examples of strong reasoning sentences and two weak ones (with explanations)"

Event 6 Prompt: "Create a mini evidence set for the prompt 'Should schools start later in the morning?' Provide 5 short evidence cards (1-2 sentences each): 2 research-style facts, 2 student experience quotes, 1 counterpoint. Include instructions for students to select any 2 cards and write a CER paragraph."

Event 7 Prompt: "Create a 6-minute peer feedback protocol for CER drafts with 3 steps, each producing written feedback. Include a mini-rubric (4 criteria, 0-2 scale)."

Event 8 Prompt: "Create a 5-minute exit ticket checking claim clarity, evidence integration, and reasoning. Provide an answer key and common wrong answers."

Event 9 Prompt: "Create 3 transfer prompts where students use CER in new contexts: science, history, and personal decision. Include a reflection prompt about evidence selection."

---

AI for Differentiation

Gagné's framework + AI enables rapid differentiation:

Same Concept, Different Levels: "Create three versions of practice problems on [topic]:

• Scaffolded version with hints and partially completed examples
• Standard version matching the lesson objectives
• Extension version with additional complexity for advanced learners"

Same Event, Different Modalities: "For the concept of [topic], create:

• A visual explanation using a diagram description
• A narrative explanation using an analogy
• A procedural explanation with step-by-step instructions"

Same Objectives, Different Contexts: "Create practice scenarios for [skill] in contexts relevant to:

• Students interested in sports
• Students interested in technology
• Students interested in arts"

---

AI-Generated Feedback at Scale

One of AI's most powerful applications: providing immediate, personalized feedback.

Teacher Prompt for AI Feedback: "You are a teaching assistant. I will paste 6 student responses to [assignment]. For each response:

1. Identify one specific strength (quote the relevant text)
2. Identify one priority improvement area
3. Write one question that pushes the student toward deeper thinking Keep feedback concise and encouraging."

Batch Pattern Analysis: "I will paste 8 student responses. Identify the 3 most common error patterns across these responses. For each pattern:

1. Name the error
2. Write a 60-second reteach script
3. Suggest a follow-up activity"

This allows teachers to provide individualized feedback and targeted reteaching efficiently.

---

Guardrails: AI That Strengthens Learning

The critical principle: AI should trigger internal cognitive work, not replace it.

Guardrail 1: First Attempt First

Before students use AI:

• Write your first attempt (even if imperfect)
• Highlight exactly where you're stuck
• Ask AI for hints, not answers

Guardrail 2: Annotation Required

When AI responds, students must:

• Annotate what they will keep, change, and why
• Include one personal example, local data, or unique justification
• Explain the AI suggestion in their own words

Guardrail 3: AI as Coach, Not Writer

Bad prompt (cognitive offloading): "Write me a CER paragraph about school start times."

Good prompt (AI as coach): "You are my writing coach. I will paste my thesis. Do NOT write my paragraph. Instead:

1. Test my thesis—ask 4 questions that challenge my assumptions
2. Suggest 3 types of evidence I could use (not specific citations)
3. Identify where my reasoning is missing and give me sentence frames to complete it Wait for my responses before continuing."

Guardrail 4: Evidence Trail

If students use AI, require:

• Their first draft (before AI)
• What they changed and why
• One paragraph explaining their reasoning

---

AI Usage Policy Framework

AI is allowed for:

• Generating examples and non-examples
• Asking better questions
• Getting feedback on YOUR draft
• Planning, checking clarity, improving structure
• Brainstorming starting points

AI is not allowed to:

• Write final submissions
• Produce citations you cannot verify
• Replace your thinking steps
• Generate content you don't understand

Evidence rule: When AI is used, attach:

1. Your first draft
2. What you changed and why
3. One paragraph explaining your reasoning

---

AI for Each Event: Quick Reference

---

The Human Element

AI enhances but cannot replace:

• Reading the room and adjusting in real-time
• Building relationships that motivate learning
• Modeling passion for the subject
• Making judgment calls about individual learners
• Creating the emotional environment for learning
• Connecting content to learners' specific lives and contexts

AI is a powerful tool. Gagné's framework ensures that tool serves learning.

---

Key Takeaways

• AI creates opportunity for rapid, scaled instructional design—with pedagogical risk if unstructured
• Gagné's framework provides scaffolding that ensures AI content supports actual learning
• AI can generate materials for every event; humans verify, customize, and deliver
• Guardrails prevent cognitive offloading: AI as coach, not replacement for thinking
• Require first attempts before AI, annotation after AI, evidence of reasoning
• AI excels at differentiation, feedback at scale, and material generation
• The human element—relationships, real-time adjustment, judgment—cannot be automated