There is a popular narrative that artificial intelligence will replace instructional designers. The argument goes like this: AI can generate content faster, cheaper, and at greater scale than humans. Therefore, the role of the instructional designer is destined for obsolescence. This narrative is not just wrong — it is dangerously backward. The truth is the opposite: AI needs instructional design far more than instructional design needs AI.
The Hidden Biases in Training Data
Every large language model is trained on a corpus of existing content. And the vast majority of educational content in the world is mediocre. It is built on assumptions that are rarely questioned: that learning is the transfer of information from instructor to learner, that completion equals competence, that engagement is a proxy for understanding, that one-size-fits-all content can serve diverse populations.
When AI models are trained on this content, they inherit these biases. They generate content that reflects the average assumptions of the training corpus — which means they generate content that is, on average, mediocre. They can produce it faster and at greater scale, but they cannot produce it better without human guidance that understands what "better" actually means.
Structure Is the Antidote
Instructional designers are trained to see structure. They understand the difference between surface content and underlying architecture. They can identify the hidden assumptions in a piece of content and question them. They know that the sequence of learning experiences matters as much as the content itself. They understand that context, audience, and constraint shape what effective learning looks like in ways that cannot be captured in a generic template.
This structural awareness is exactly what AI lacks. AI can generate plausible-sounding content that follows surface patterns. But it cannot judge whether the structure beneath that content actually supports learning. It cannot identify the missing reflection prompt that would turn passive consumption into active processing. It cannot recognize when a sequence is building cognitive load in the wrong order. It cannot sense when the content is culturally inappropriate for the target audience.
The Partnership, Not the Replacement
The most productive relationship between AI and instructional design is not replacement — it is amplification. AI handles the generative work that designers find tedious: drafting initial content outlines, producing first-pass materials, generating variations for different audiences. Designers handle the judgment work that AI cannot do: evaluating whether the structure works, identifying missing pieces, adapting to context, and making the subtle decisions that turn generic content into effective learning.
At Mekalin, we use AI as a support layer in our design process. It generates frameworks that our designers then evaluate, modify, and refine. It surfaces risks and dependencies that might otherwise be missed. It handles repetitive tasks so that designers can focus on the high-judgment work that only humans can do. The result is not AI-designed learning — it is AI-assisted, human-designed learning. And that is a fundamentally different thing.
The Future Is Augmented, Not Automated
The organizations that will thrive in the age of AI-powered learning are not the ones that replace their designers with AI. They are the ones that equip their designers with AI — giving them tools that extend their capabilities while preserving the judgment, context-awareness, and structural thinking that makes instructional design valuable in the first place.
AI does not make instructional designers obsolete. It makes their skills more essential than ever.
