Instructional Design in the AI Era

Over 80% of instructional designers reported using AI in 2025, and it’s safe to assume that number has grown this year. (At Six Red Marbles, all our designers use AI daily.) They’re using AI for everything from proposing guidelines for content development to formatting documents and creating visuals. At colleges and universities, instructional designers are on the front lines of AI innovation in course design, and that innovation has one main goal: speed. More than anything else, the increased speed will reshape instructional design work.

AI Is Accelerating Instructional Design

At first blush, increased speed seems like a good thing. Courses can be generated more quickly. Polished drafts make review cycles easier. Accessible supports like transcripts and alt text are easier than ever to create. But speed brings its own complications, especially for higher education.

Brevity is the new differentiator

AI is verbose. (If you’ve figured out how to get your LLM to write a short sentence, please let me know.) Just because you can generate an incredibly comprehensive overview of a topic in seconds doesn’t mean that your learners will actually read that overview. In all likelihood, students will put large blocks of AI-generated text back into their own AI to summarize.

In a world where being verbose is as easy as hitting enter, brevity may be the new differentiator. This does not mean that English majors should stop reading Faulkner. But it does mean that we should think twice before using AI to generate a 40-page reading and study guide for Chapter 1 of The Sound and the Fury. Students are much more likely to engage with a single well-thought-out question on a discussion board or respond to a quick video that motivates them to keep slogging through Benjy’s stream of consciousness.

Programs need to highlight what makes them different

When AI is leveraged heavily in drafting and design, what it creates will align to whatever every other publicly available program is doing. (We may think of AI as a content generator, but it’s really just aggregating information.) This presents a new challenge for institutions that still need to make their programs stand out.

Institutions should look for opportunities to weave their mission, vision, and values into course design. Courses should reflect the expertise and unique viewpoints of faculty, as well as the quirky, idiosyncratic approaches to presenting material that can only come from humans. The most differentiated courses will have content that was the product of human brainstorming, not machine synthesis.

Content management becomes a bigger problem

Every shortcut comes with a catch. Faster creation means more content to maintain. All courses and programs need periodic updates and refreshes in addition to day-to-day troubleshooting. While a small ID team may suddenly find themselves able to create dozens of courses in the time it previously took them to make two, the long-term implications of this choice are that they will have significantly more content to keep up-to-date.

So what is a reasonable workload for an instructional design team with AI? If instructional designers can create more content more quickly, it would seem logical to give them more courses to work on, especially if all they’re doing is prompting and quickly reviewing AI-generated content before passing it along to a faculty SME. But institutions should beware of falling into the trap that AI means everyone should work more quickly. Every time an instructional designer switches tasks, there is a mental toll. Even pre-AI, task-switching was estimated to cost up to 40% in lost productivity. Even if the human ID is reviewing more content than they’re creating, they can only juggle so many projects simultaneously.

The Human Take

It may be tempting to outsource these very problems to AI. Isn’t AI’s promise to solve our complex problems and provide comprehensive, actionable results faster than we ever could do ourselves?

For fun, I asked ChatGPT what we should do. After 17 minutes of pro-thinking, it returned an eight-step guide that it promised it could turn into a pithy one-pager. Which kind of encapsulates the problem.

So here’s the human take I generated while I was waiting on ChatGPT.

Before outsourcing a problem or task to AI, it’s worth pausing and thinking about whether it will offer any benefit. If you’re outsourcing something to AI, it’s safe to assume others are too. When every institution can generate an AI-created data science program and students can prompt AI to teach them Python, knowing when to leverage AI—and when not to—becomes one of the biggest differentiators for an institution. For now, that still takes a human touch.

Jocelyn Wright, Ph.D., is Director of Community & Market Development at Six Red Marbles. She writes a monthly higher ed newsletter, The Close Read, and hosts a community of practice called The Credit Hour. The team at Six Red Marbles would love to stay in touch about all that’s facing higher education. Reach out if you’d like to continue the conversation.

Six Red Marbles (SRM) partners with colleges and universities to design and deliver high-quality online, continuing, and professional learning experiences. We support institutions across the full lifecycle of program development from semester and course preparation to credential design, faculty-ready materials, and product development workshops. Our work blends human-centered instructional design with innovative, scalable solutions to help institutions move faster while maintaining academic integrity, consistency, and learner impact. SRM collaborates closely with academic and administrative teams to modernize learning in ways that are practical, sustainable, and aligned with institutional mission.