Most IDs bring AI into development. Scripts. Storyboards. Slide copy. That’s fine, but you’re leaving the most valuable use case on the table.
The real leverage is earlier. In Design, when you’re still deciding what the course should be.
AI doesn’t replace your instructional judgment. It pressure-tests it.
Here’s how to use it at the strategy level, before a single slide gets built.
The Problem with Designing Alone
When you’re deep in a project, you’re also deep in assumptions. You know the SME’s perspective. You know the stakeholder ask. You’ve already started thinking in a format: probably eLearning, probably 30 minutes, probably the same structure you used last time.
AI can break that pattern, not by giving you the answer, but by asking the questions you stopped asking.
Three Ways to Use AI for Instructional Strategy
1. Challenge your modality default
Before you commit to a format, ask AI to push back on it. Give it the context: the audience, the performance gap, the constraints. Then ask it to argue for a different approach.
You might still land on eLearning. But you’ll land there with a reason, not a habit.
2. Map your learning objectives to a real sequence
Most IDs write objectives and then sequence content by topic. AI can help you sequence by cognitive load instead, surfacing what learners need to know first before they can process what comes next.
Paste in your objectives and ask AI to suggest a scaffolded sequence, then tell you why. Interrogate the logic. You’ll often find a better order.
3. Stress-test your structure before you storyboard
You’ve got a course outline. Before you develop it, give it to AI and ask: “What’s missing? What would confuse a learner? Where are the logic gaps?”
Think of it as a pre-mortem on your design: cheap to run, often catches real problems.
Prompt of the Week
Use this at the start of the Design phase, once you have your objectives and some audience context:
You are an experienced instructional designer reviewing a course structure before development begins.
Here is the context:
- Audience: [describe learners, role, experience level, motivation]
- Performance gap: [what they're not doing well and why it matters]
- Learning objectives: [paste them here]
- Proposed format: [eLearning / ILT / blended / job aid / etc.]
- Constraints: [time, budget, LMS, etc.]
Please do three things:
1. Suggest a sequencing logic for the objectives, ordered by cognitive scaffolding, not topic grouping. Explain your reasoning.
2. Flag any objectives that seem misaligned with the performance gap.
3. Challenge the proposed format: is it the right fit for this audience and gap? If not, what would you recommend and why?Run this before you touch a storyboard. The pushback is worth it.
If you’ve been treating AI as a content generator, try using it as a design reviewer this week. It won’t replace the judgment call (that’s still yours), but it’ll make sure that call is deliberate.
If you want a structured library of prompts built specifically for the ADDIE phases, that’s exactly what we build inside the AiDDIE Vault. Come check it out at aiddie.co.
— Gus
P.S. Hit reply if you try that prompt. I’d love to know what it surfaces for you.