The duty invites you once again to the back roads of university life. A proposal that is both scholarly and intimate, to be picked up all summer long like a postcard. Today, we are interested in what AI can bring to those who teach.
For many people, artificial intelligence (AI) in education is all about ChatGPT and the ability to use it to cheat. However, according to a recent study by KPMG in Canada, while 52% of Canadian students aged 18 and over use generative AI to help them with their schoolwork, 60% of them feel they are cheating by doing so.
Canadian universities are not quite ready to ban these tools, however. Indeed, generative AI is a complex tool that, beyond cheating, has the potential to promote learning by facilitating the design of educational materials, which is our area of expertise.
Let’s first look at why new technologies are off-putting to many people.
History shows us that any technological innovation meets with some resistance at the beginning. Let us recall that in 1895, one of the Lumière brothers’ first films frightened the spectators, who fled for fear of being hit by the locomotive entering the station.
Although it is now considered a golden age, in the 1950s many considered television a cultural wasteland. When calculators first appeared in the 1970s, teachers tried to ban them for fear that students would lose their mastery of mathematics.
Many academics denigrated Wikipedia when it first appeared in the early 2000s. Less than a decade later, during a meeting of the Concordia University senate, the then president acknowledged in a comment the authority that the tool enjoys.
In short, history repeats itself. When a new technology comes out, it usually meets with some resistance from people who often know very little about it. However, the more people try this technology and apply it in their work and everyday life, the more it ceases to be an abstract and potential threat and becomes a practical and useful tool.
This is exactly what is happening in the field of education. Just 19 months after the launch of ChatGPT, education specialists are starting to use this technology to optimize their work, especially in their area of specialization, instructional design.
Instructional design involves developing the teaching materials and approach in such a way that they enable the acquisition of a specific set of skills and knowledge.
Instructional designers use science, arts, and technology across the curriculum—from kindergarten to college—as well as in workplace training, continuing education, and community learning. We integrate technology into our work, but only when it is useful for educational purposes, and we always keep in mind the people who will use the technology.
Although some AI tools offer the possibility of developing entire courses, given the diversity of the student population, the complexity of the subjects to be mastered and the current state of development of AI, we do not fear that artificial intelligence will soon replace instructional designers.
However, we are discovering concrete ways to help develop even more effective teaching tools.
Specific applications
Although generative AI can be used to analyze, design, develop, implement, and evaluate instructional materials, we will focus on how it can be used to address problems specific to the analysis and design stages.
Overabundance of material. Instructional designers often have only complex and poorly written documents from which to draw curriculum material and teaching elements. Educational researchers have formulated queries (or prompt) that allow ChatGPT to be used to find, summarize and synthesize reference documentation and thus avoid the excess of information associated with preliminary research.
Learning context. It often happens that an instructional designer must deal with a subject that he or she knows little about or that seems reserved for insiders. By means of judiciously formulated queries, it is possible to learn about the subject and obtain a context that even specialists in the field would struggle to provide.
Access to learners. Instructional designers often have little direct interaction with the learners who benefit from the fruits of their labor. So researchers have formulated queries that can be used in ChatGPT to create personas, which designers use to understand their target learners.
Writing goals. One of the first tasks an instructional designer tackles is developing observable and measurable objectives that will serve as a framework for the program. Tools like Ontario-based Contact Nord’s AI Teaching Assistant and the University of Saskatchewan’s SMARTIE help designers write such objectives.
Writing assessment questions. Immediately after writing objectives, designers prepare assessments that measure how well learners have mastered those objectives. ExplainIT, developed by researchers at Indiana State University and North Carolina State University, helps teachers assess understanding of material by asking students open-ended questions and immediately responding to their responses in class. AI Teaching Assistant and SMARTIE can write exam questions and prepare rubrics to assess responses to open-ended questions. AI-powered tools perform similar tasks in the workplace.
Additionally, because AI can write assessment questions more easily than humans, teachers are more likely to design new exams each term, which can reduce the risk of cheating. (Many exams are sold online after they’ve been used. When teachers reuse exams, they increase the opportunity for cheating.)
Teaching strategies. Designers then create learning materials that meet the objectives and prepare learners to take the exam. Designers have begun to use AI to make important decisions about which strategies to use. They can now determine whether to deliver a course online or in a classroom, whether to give a presentation followed by exercises or whether to focus on “exploratory” learning, and even calculate how to break down content into lessons to make it easier to absorb the material.
All of these tasks take a lot of time, which prevents many designers from doing them fully efficiently. This was true until the advent of AI. Artificial intelligence does not reduce the work required to design instructional materials. It does, however, optimize the time spent on it.