A reflection on Weller’s Learning Objects chapter in 25 Years of Ed Tech: When technology combines reusable content with adaptable personalized learning.
Reading Weller’s 25 Years of Ed Tech, I was surprised by how immediately the concept of learning objects made sense to me. As a designer and developer, I am accustomed to thinking in terms of modular, reusable components that make products and systems easier to maintain, scale, and adapt. Weller describes learning objects similarly: educational content broken into reusable pieces that can be assembled into different learning experiences (2020, p. 50). My first reaction was therefore: Of course. Why wouldn’t we build learning this way?
Chapter 7, and particularly the podcast discussion, made me question that instinct. One problem with learning objects was that the more reusable an object became, the more context had to be stripped away. Yet context is often what makes learning meaningful (Weller, 2020, p. 52). The supplementary podcast also highlights how concerns about intellectual property, attribution, and the perceived commodification of teaching contributed to resistance to sharing (Pasquini, 2020). This resonated with my experience in post-secondary education. When I began teaching in 2018, I was surprised that faculty did not routinely share course materials. Coming from an agency environment, I was used to designing systems so that someone else could pick up the work if I was unavailable. During the pandemic, these concerns became even more visible as faculty worried that online materials could turn their expertise into easily transferable commodities.
I still believe strongly in reuse. However, I am increasingly interested in the relationship between technology-mediated content and learning. A reused learning object can deliver information, but it cannot necessarily reproduce the relationships, context, feedback, facilitation, and community surrounding it.
What I find particularly interesting is how familiar the promise sounds today. With AI capable of manufacturing learning objects at enormous scale, the cost of producing content drops dramatically. More importantly, AI may address one of the central problems Weller identifies: context. Rather than creating a highly contextualized object that is difficult to reuse, AI can potentially adapt a generic object to an individual learner by personalizing content, adjusting learning pace, and providing real-time feedback (Tan et al., 2025). In this sense, AI could make the learning-object vision more achievable than its creators imagined. Consequently, if AI can generate the objects and adapt them to the learner, what remains uniquely human about teaching?
Yet personalization is not the same as context, feedback is not the same as dialogue, and AI alone is not a Community of Inquiry. Tan et al. (2025) identify privacy concerns and institutional support as significant barriers to adopting adaptive learning platforms. Perhaps the history of learning objects reminds us that making content cheaper and faster does not necessarily make learning better. The challenge may ultimately be less about what the technology can do than whether it creates a learning experience that faculty and students value enough to adopt.
Footnote: AI was used to improve the language in the final draft of this reflection. All ideas are authentically mine.
References
Pasquini, L. (Host). (2020, December 24). Between the chapters: Learning objects [Audio podcast episode]. In 25 years of Ed Tech: The serialized audio version. OpenEd. https://25years.opened.ca/2020/12/26/between-the-chapters-learning-objects/
Tan, Le Y., Hu, S., Yeo, D. J., & Cheong, K. H. (2025). Artificial intelligence-enabled adaptive learning platforms: A review. Computers and Education: Artificial Intelligence, 9, 100429. https://doi.org/10.1016/j.caeai.2025.100429
Weller, M. (2020). 25 Years of Ed Tech | Athabasca University Press. Aupress.Ca. https://doi.org/10.15215/aupress/9781771993050.01
Great post, Claire! Your reflection on the “Lego block” approach to learning objects really captures the long-standing debate between technical scalability and pedagogical effectiveness. While Weller highlights the loss of context as the primary structural flaw and central paradox of the learning object movement, I actually view this a bit differently. I don’t see modularity and context as mutually exclusive. Instead, effective learning design functions as a hybrid: you use standardized “Lego blocks” for core concepts, and then layer on specific, personalized context depending on the delivery format. For instance, in my own work, a foundational visual like an Iceberg Model of Knowledge can serve as a reusable core asset. The asset itself stays modular, but its context shifts completely depending on whether it is introduced in a blog post, a formal course module, or a interactive discussion. The object provides the explicit framework, while the surrounding facilitation, medium, and prompt supply the localized meaning. Your point about AI potentially automating that contextual layer is a fascinating next step. As long as we recognize that the core block and the contextual wrapper work together, modular design doesn’t have to mean stripping away meaningful learning.
Hi Zubeida,
Thank you for reading my post and for pushing my thinking a little further. Your point that reuse and context aren’t necessarily mutually exclusive makes a lot of sense. A learning object can remain modular and reusable, while the instructor provides the contextual layer that connects it to the rest of the course and the learner’s experience.
I do still wonder, though, whether this was part of the challenge with the original learning-object vision. As a designer, I find it hard to imagine that a course assembled from dozens of discrete objects would necessarily have the same coherence and intentional flow as one designed from scratch. Perhaps the question is how much contextual work is required to make a reusable object meaningful, and whether that work undermines some of the efficiency modularity promises.
I also found Weller’s description of learning objects as a “successful failure” interesting (2020, p. 54), particularly because they helped contribute to the evolution of OER. Many contemporary OERs seem less granular than the original learning objects were intended to be. Perhaps their lasting contribution was less the Lego-block model itself and more the idea that educational resources could be shared, adapted, and reused.
And now AI potentially adds another twist! I am guarded when it comes to AI. While I recognize its potential, I still believe that transformative learning fundamentally requires human connection: at least two humans in a (digital) room. Maybe the human connection is the context that no LO nor AI can replicate.
Claire
Hi Claire,
Excellent post! Your follow-up comment to Zubeida about human connection being the context that cannot be replicated is central to my own understandings of why AI should never be the sole teacher in a learning environment. I have seen a few schools advertising a personalized learning experience with AI facilitators as the core structure of the program. The world of Social Media (Instagram, Facebook etc.) has provided a precursor to these types of environments in that an algorithm curates these objects (reels, posts, etc. –calling them LO would be a stretch) for the individual’s viewing. What I’ve noticed is the innate connection piece that comes to the fore in this interaction. When a particular post or reel lands, there is an overwhelming urge to circulate and share. Not only that, there is often dialogue associated with it that embodies the co-construction of knowledge. This example depicts the desire for human connection across technology platforms, demonstrating the sociocultural aspect necessary within a learning environment and for learning to occur. So while AI may gather LO for consideration along your personalized learning path, I believe the link to constructivism is where this model would fail to resonate with the learner.
In my view, it almost seems that AI goes back to an authoritarian approach to learning, centred on instructivism. Viewing learning objects as one part of a learning program is crucial, as I believe that deep learning does not occur in isolation. Based on my own experience as a teacher for 20 years, I also believe that learners make meaning through hands-on, experiential opportunities that would be hard to replicate in even the most well designed learning object. This exemplifies why I believe they should serve as one aspect of a learning program.
Finally, I feel like Weller (2020) minimizes the importance of LO in his retrospective account. His assertion that learning objects failed feels overstated and inaccurate. I believe learning objects have certainly shifted over time but still play an important role in learning opportunities. I have not attended many professional development sessions without a powerpoint presentation, which I would argue is an LO, and a heavily utilized one at that.
Talk soon,
Jenna
Hi Jenna,
thank you for your reply. II agree that Weller may have overstated the “failure” of learning objects. Your example of PowerPoint also made me think about how broadly we might define a learning object.
I also share your concern about where AI-driven personalization could take us. It’s sad that we are entering an age where faculty, designers, developers, and many other professions increasingly have to justify their value in the face of “efficiency” arguments. Technological advances have always changed jobs and created new ones, but I wonder if the current pace of change is making it harder for organizations to see the long-term costs alongside the immediate efficiencies. We can measure the cost of replacing a human task fairly easily; measuring the value of the relationships, creativity, mentorship, and knowledge-sharing that disappear with it is much harder.
Claire
Hi Claire,
I really appreciated your thinking around learning objects and commodification of use. When I entered the post secondary field I was also surprised by the hesitancy to share materials, when I worked as a floor nurse I shared ideas and expertise constantly with my peers. In the nursing department at the college I work at we still share materials and content readily but this isn’t the case for the college at large and I can understand why. Recently management at the college has asked us to make generic course shells, lesson plans, and other content so that anyone with content expertise could teach the material, we has a department have pushed back against this as we see it as a way that could be used to eliminate positions.
To your points around AI I don’t think I can agree, my understanding of AI is that it feeds people what they want to hear and even more context would be needed to combat AI’s propensity for this. I also don’t see AI at a place where it could teach, I have my students to a project where they have to research a topic and we find that AI frequently hallucinates information and data, I only really see it being a useful tool if you are already a topic expert and and quickly weed out the erroneous information. I used AI quite heavily just over a year ago in my work and studies, mostly for improving my writing and to bounce ideas off of, I’m trying very hard to not use it anymore as I worry about environmental impacts and AI’s place in the world at this time… hopefully my writing doesn’t suffer to much!
Thank you for the thoughtful post, look forward to hearing more!
Callan
Hi Callan,
We’ve also been asked to prepare and share our courses so they can easily be handed over. There’s another tension here that I find interesting: administration seems to want faculty who are experts in their field while also expecting them to teach the entire breadth of that field. I teach Web Design and Development, for example, but in industry you rarely find a designer who also specializes in back-end development, and vice versa.
I understand your decision to step back from AI, particularly given the environmental concerns. I’m still experimenting with it and trying to understand where it is genuinely useful. I’ve actually switched my default search engine to DuckDuckGo’s no-AI option so that I’m only using AI when I deliberately choose to.
And just to clarify my original post, I definitely wasn’t suggesting that AI can replace human teachers! I was thinking more about the possibility of AI automating some of the less meaningful work around learning objects, such as generating metadata, while leaving the human work of teaching intact.