ChatGPT and the Future of Digital Learning
Justin is Assistant Director of District Programming at Michigan Virtual
Will AI-powered tools like ChatGPT transform education as we know it? The recent posts here on the DLC blog make great cases both for and against the potential for transformational change. I think it’s still too early to tell, as the full capabilities of the technology and how well they align to the core processes of learning are still unknown. I do think, however, that it’s not too soon to speculate on how this tool and others like it might impact digital and online learning specifically. As the use of ChatGPT begins to spread in educational circles, certain aspects of the discussion around it are worth exploring further to see how they relate to digital and online learning.
Probably the most prevalent topic in coverage and discussion of ChatGPT thus far has been its use by students to produce text for class assignments. Commentators are taking sides on whether to ban the tool when using school-provided devices, and tools like GPTZero are garnering attention for their efforts at detecting AI-produced text to suss out instances of plagiarism. In a recent episode of the New York Times podcast Hard Fork, a high school English teacher gives her perspective on how and when both students and teachers should lean on the tool for assistance. She mentions that her department is already discussing ways to revise the writing and assessment cycle, by having students write smaller chunks of text in person in the classroom for assessment and revision, in order to better control their access to ChatGPT. A similar argument is made in a recent blog post by Matthew Tower, contending that we should bring back Blue Books and require handwritten essays. These practices would be a pretty big deviation from the more recent practice of having students write lengthier assignments at home, and it raises an obvious question for me: how would this process serve (or not serve) students who learn online or outside of a brick-and-mortar classroom? As these AI-powered tools become more capable, requiring any assessed work to be done in a tightly controlled environment that prohibits them seems like an untenable solution.
When it comes to the provision of content and instruction, online/hybrid teachers and instructional designers could certainly see some immediate effects of ChatGPT’s use. Unit and lesson outlines, learning activities, and assessment criteria can all be produced with ChatGPT to a decent level of adequacy, especially when using specific content and learning standards within the prompts. As another recent blog post illustrates, it’s easy to envision how using it could accelerate the process of online course design and revision, allowing online providers to more easily offer an extensive and up-to-date catalog. It could also be used to streamline personalized learning efforts, as teachers and instructional designers could theoretically more quickly produce content and assessments that could better align with students’ interests or strengths. Similarly, teachers and instructional designers could use the tool to help with coding efforts to design more engaging and interactive digital learning experiences.
Another thing to consider will be how quickly providers of other online tools adopt the underlying technology that powers ChatGPT. It’s likely that the foundational process of web search will change, allowing users to get more personalized answers to their queries when visiting Google, Bing, or perhaps even a disruptive upstart competitor. Likewise, tools like online word processors and slideshow creators will probably soon have GPT functionality integrated to facilitate the writing and presentation creating processes. Simply banning ChatGPT will be irrelevant if students are using commonplace online tools with the same underlying capabilities. It’s also not hard to see learning management systems potentially jumping on board, an evolution of the work that’s already focusing on AI-dependent nudges for changing student behavior or AI-powered recommendation systems to personalize learning. LMS providers would no doubt explore the possibility of adding GPT-powered features in an attempt to capture a larger share of the market if such features can prove valuable for providers of online learning.
Taken together, all of these potential developments could drastically change how digital learning experiences are created, delivered, and perceived. Finding the sweet spot in the tension between tight control and loose facilitation of learning experiences will be key to preserving the flexibility and other affordances of digital learning as these technologies become more prevalent. What remains to be seen, however, is whether ChatGPT and tools like it can bolster the foundational elements of learning, whether in-person or digital: relationships and human connection.