From the outset, one of the key promises of AI in education has been relieving teachers of repetitive, redundant tasks so they can channel more energy and time on meaningful interactions with learners. Canvas’s new partnership with OpenAI introduces features, such as rubric generation and grading automation.
But what captivated me even more is what lies beyond the rubric generation and grading automation. Recently, I shared on LinkedIn that Claude now supports Canvas LTI, which allows learners to access the generative artificial intelligence (GenAI) tool directly from within their Canvas courses.
In the post, I mentioned that as time goes by, we’re going to see more innovative integrations of GenAI tools with edtech solutions. Small wonder, this week, OpenAI, Anthropic’s competitor, has announced its own strategic partnership with Instructure, the creators of Canvas Learning Management System, that brings new AI capabilities to the LMS.
At the heart of this partnership lies a vital question which has echoed across the world since the launch of ChatGPT in 2022: How can we ensure effective incorporation of AI into learning without compromising educational integrity, especially within institutions of higher learning?
This partnership aims to help answer this question through an innovative assignment delivery approach. By seamlessly embedding AI into learning experiences, not as hidden external tools, but as active agents of learning, it can help dispel the widely held notion that learners primarily use the tools to cheat. Instead, integrating and openly exposing learners to these tools enhances their learning experiences while at the same time fostering trust between them and their teachers.
Now, Canvas users can have conversational assignments, which are embedded in AI-driven personas. This turns assignments almost completely on their heads by moving away from the passive reception of information and exploration of ideas by learners to more active, iterative, and responsive interactions with AI-powered chatbots.
Even more importantly, while learners converse with AI on assignments, teachers review the back-and-forths between learners and the chatbots, hence revealing information about the thought process of the learner. Additionally, teachers can develop rubrics that cover important aspects of the assignment, such as proper and relevant mentioning of key concepts and personalities, sources mentioned, and critiquing of the AI-generated outputs.
Furthermore, learners can’t take shortcuts by simply prompting the AI to “give me answers”. In such a scenario, the AI will push back, requiring the learners to critically and meaningfully engage with the assignment. Even without the pushback, the teacher would see, when reviewing learners’ submissions, that a student prompted the chatbot to simply ‘vomit’ the answer.
For example, in the video that has been released by OpenAI, an assignment is created on an economics policy course. The teacher instructs learners to take a position to support or oppose government intervention in the economy, while drawing from John Maynard Keynes’s economic theories.
Shiren Vijiasingam, Chief Product Officer at Instructure, asserted in a press release that these AI-powered assignments within Canvas “allow[s] students to have rich, casual, and interactive conversations in a ChatGPT-like environment they are already used to using.”
“In the process,” he added, “[students] create visible learning evidence that teachers can confidently use, as it’s mapped to the learning objectives, rubrics, and skills defined by the teacher.”
Essentially, the OpenAI and Instructure partnership reveals how AI can complement learning transparently and ethically, while enhancing learning experiences and practically redefining the role of educators.
Again, we are still going to see more innovation in AI integration in edtech solutions, and eventually, transformative approaches to education. For example, in research methods or academic research, in the literature review, I anticipate revolutionary approaches that embrace the use of AI to suggest and review sources.
Educators will encourage learners to use AI in identifying patterns to avoid mistakes that have been made by other researchers. Additionally, it will shorten the life cycle of research projects by revealing new innovative techniques that researchers can use to benefit from other researchers’ experiences.
Africa, we still have a long way to go on this journey. While many institutions of higher learning are at the nascent stages of introducing open distance e-learning modes of study, where LMSs, such as Canvas, come in handy, the infrastructural and pedagogical concerns remain primary barriers.
Lastly, learning experience specialists must continue to rigorously interrogate these integrations to ensure that they are anchored on the science of learning principles and that the tech corporations, such as OpenAI and Anthropic, are constantly held accountable to ensure efficiency and effectiveness.
AI hallucination remains a significant pedagogical challenge. This must be addressed in educational settings. Otherwise, these technologies can turn out to be a curse rather than a blessing.
1 comment
Comments are closed.