AI company CEOs are claiming that artificial intelligence
will replace workers (Cutter & Zimmerman, 2025) based on the vast amount of
content quickly created by AI. However, faster does not necessarily entail
better. When one needs brain surgery, it
is probably not wise to go with the surgeon who claims to perform the procedure
faster than anyone else with no attention to the survivability rate. When stakes are high, we value accuracy. Sacrificing
a marginal amount of time becomes an acceptable trade.
Online learning took years of research to confirm its legitimacy, and in 2009 the Department of Education confirmed that not only were online courses comparable, but were in some cases better than, traditional courses (Means, et al., 2009). Once the pandemic necessitated a move to online instruction, a vast number of academics moving to online instruction frequently resulted in best practices being ignored in favor of speed by those new to the online modality (Farrell, 2022). Years later, the effects on online instruction without proper instructional design is still not determined.
The introduction of AI provided tools has produced more AI
generated text than the combined output of human generated text since the Guttenberg
printing press (Vincent, 2021). This
vast amount of content is hawked as a good thing. However, quality isn’t measured in terabytes.
When writing computer programming code, AI code often contains bugs and
introduces security risks (Perry et al, 2023). Alarmingly, while programmers predicted that
AI tools can reduce their completion time by at least 20%, studies reveal the
tools slowed down completion by 19% (Becker, et al, 2025). This significant
difference illustrates that if one is enamored by a quick an easy tool it could
lead to a sacrifice in quality that will impede the end result. Simply having a
tool that can produce vast amount of content does not vouch for the quality of that
content.
How does this apply to online learning? Over twenty years of research molded an online design approach that focuses on positive learning experiences (Wasson & Kirschner, 2020) where courses are designed to be accessible to all students, follow a Universal Design for Learning approach (Dell 2015), and focusing on complex learning through engaging designs instead of merely presenting content (Wasson & Kirschner, 2020). A few features that are expected in these online designs include:
- A focus on the learner
- Prototyping designs
- Designs that accommodate multiple learning strategies (Rose & Meyer, 2002)
- Iterative development incorporating student feedback (Adnan & Rizhaupt, 2018)
- Aligning learning outcomes to assessments (Ni She, et al, 2021)
Being able to develop online learning environments that apply best practices defined by applied research in learning is critical for creating courses that promote student success. While AI may be able to quickly produce content, would it be able to create discussion activities for adult learners that:
- Promote active learning
- Provide open-ended questions that explore and apply concepts
- Encourage learners to apply their real-world experience to the content
- Avoid soliciting facts that close off conversations
- Apply scaffolds that ensure inclusive learning practices
With the vast growth of the AI industry, faculty access to
AI, and economic issues facing schools, it is quite possible that AI will quickly
provide shoddy content at the cost of the learners.
Case Study: Brightspace LUMI AI
- test questions,
- discussion questions,
- assignment ideas, and
- module summaries.
This can be done quickly and relatively easily. Impressively,
users can even set the level of Bloom’s taxonomy for many of the outputs. Being
integrated within Brightspace allows for a common platform for a school to use
that will save money and training time.
Its one-press button is also convenient for the faculty.
Will it replace developers and instructional designers? Like the claims of many selling AI tools, they
are exaggerated. When using AI to
develop questions for online discussions, I noticed a disturbing trend. The questions were similar to a new teacher
who has no knowledge of best practices in instructional design.
With each generated output, the discussion questions lacked
instructional design principles for online learning. Presumably developers will
edit the material and make necessary additions or changes. In reality, overworked instructors are not
incentivized to do so. And this is the
crux of the problem with AI. It
encourages the quick and easy solution and contributes to ignoring specialists.
AI slop is quick and easy; however, it does not promote student learning (Weller,
2024).
Below is a LUMI Pro AI training guide:
The Lesson
AI is a tool – not a solution. If you are going to use AI, be sure to edit and apply best practices in instructional design for learning. For example, online discussion questions should be written with:
- Open ended questions
- Providing follow-up questions and time for reflection
- Linked to authentic life experiences
- Encourage sharing references and promoting an academic conversation.
While there is nothing wrong with using AI to help inspire you when developing courses, you will probably note that the best courses have inspiring and engaging learning activities that are more than just a vague question for a standard online forum. To be fair to LUMI, Brightspace has designed it such that instructors must review/edit the generative results before they are applied.
The AI Apocalypse - Wow?
Replacing experts with stochastic parrots is probably a
recipe for disaster. The hallmark of good
college is its instruction. AI is a tool
and it is important to not be enamored by its novelty. Instead, learn to work with it and know its
limitation. Be sure to vet all generative
AI material. Be warned: You may be like the programmers who think the
AI tools are increasing their speed, when in actuality the tool has become an albatross.
Since instructional design is a detailed field, it is prudent
to understand best practices in online course development to assist your
assessment of the AI content. Most
teaching and learning centers are filled with instructional designers who are
more than happy to share the knowledge with faculty. While the may dispel the razzle-dazzle of
some AI tools, it will contribute to developing courses that increase student success.
References
Becker, J., Rush, J., Barnes, E. & D. Rein (2025) Measuring the Impact of Early-2025 AI
on Experienced Open-Source Developer Productivity. arXiv.
Cutter, C & H. Zimmerman (2025) CEOs
Start Saying the Quiet Part Out Loud: AI Will Wipe Out Jobs. The Wall Street Journal. July 2.
Dell, C. A., Dell, T., & Blackwell, T. (2015). Applying Universal Design for
Learning in Online Courses: Pedagogical and Practical Considerations. Journal
of Educators Online, 12(2), 166–192.
Farrell, O. (2022) Learning
Design in the Time of COVID-19: The Digital Learning Design Unit Story.
Open Praxis.
Hodges, C., Moore, S., Lockee, B., Trust, T., & Bond, A.
(2020, March 27). The
Difference Between Emergency Remote Teaching and Online Learning. Educause
Review.
Means, B, et. al (2009) Evaluation
of Evidence-Based Practices in Online Learning: A Meta-Analysis and Review of Online Learning Studies. US Department
of Education.
Ní Shé, C., Farrell, O., Brunton, J., & Costello, E.
(2021). Integrating design
thinking into instructional design: The #OpenTeach case study. Australasian
Journal of Educational Technology, 33–52.
Perry, N, Srivatava, M, Kumar, D, and D, Boneh (2023) Do Users Write
More Insecure Code with AI Assistants?. In Proceedings of the 2023 ACM SIGSAC Conference on Computer and
Communications Security (CCS ’23), November 26–30, 2023, Copenhagen,
Denmark. ACM, New York, NY, USA.
Rose, D. & A. Meyer (2002). Teaching every
student in the digital age: universal design for learning. Alexandria, VA:
Association for Supervision and Curriculum Development.
Vincent, J (2021) OpenAI’s
text-generating system GPT-3 is now spewing out 4.5 billion words a day. The Verge. Mar 29.
Wasson, B., & Kirschner, P. A. (2020). Learning Design: European
Approaches. TechTrends, 64, 815–827.
Weller, M (2024) 30+
Years of Ed Tech – 2024: AI Slop. The
Ed Techie.