Teaching AI Competencies: Lessons from Academics Incorporating AI in the Classroom
In the ever-evolving landscape of adult education, the ascent of artificial intelligence (AI) represents a seismic shift poised to reshape knowledge work. A recent McKinsey report aptly states, “Generative AI can substantially increase labor productivity across the economy, but that will require investments to support workers as they shift work activities or change jobs.” Education institutions play a critical role in preparing workers for these changes.
As part of the AI in Adult Education series, the team at Ribbon Education brought together voices from diverse education institutions—Asim Ali, Executive Director of Center for Enhancement of Teaching & Learning at Auburn University; Kaari Casey, VP of Academic Operations at Campus; and Jessica Mitsch Homes, Co-Founder and CEO at Momentum Learn—to discuss the topic of “Teaching AI Competencies.” They share their takeaways and examples of how they infuse AI into teaching and learning.
We [education providers] play a role in helping learners understand how AI is going to impact their work.
Asim: It’s crucial for those interacting with learners, whether through certification or credentialing programs, to ensure learners walk away with a clear understanding of how AI will affect their work.
For example, where I used to assign an essay I now ask students to interact with generative AI tools that have been vetted by the university. It’s helpful for students, after using these tools, to reflect. Where do they see this stuff going? How do they see this impacting their work? I think it’s very important, as they go through the rest of their coursework, to have this frame of mind.
Kaari: It’s on us to help industry understand what they’re looking for and know how to assess which candidates are meeting that criteria rather than the other way away. And we need to equip our students with the ability to articulate what they’ve learned and translate it into the skills that can create a positive business impact.
Jessica: We work backwards from what employers are currently doing. This spring, we made calls to 122 companies to ask them how they’re using this technology. We use that information to prepare our students for that reality. There is a lot of hype around AI right now and pressure from the boardroom to figure it out. But that’s not necessarily coming from the practitioners doing the work. So, we play a big role in helping companies define what it really means at this moment.
Allow students to use AI tools only once they have strong foundational skills to effectively utilize the tools.
Jessica: These AI tools speak with such authority, and we know they are not always right. So, we want our beginner learners to develop that muscle memory on their own, to really go through the process of writing and explaining their own code and not having this other entity interact and potentially take them down the wrong path. That said, we know they are likely going to have access to it on the job, so once they get to a certain proficiency level, we do expose our students to these tools and teach them how to leverage them.
Kaari: We take a two-pronged approach in which the foundational skills must be rock solid. Our students need to have good reading comprehension. They need to know how to think critically to spot hallucinations. So, we take the initial approach of ensuring the fundamental skills exist—so maybe less ChatGPT usage in our early-stage courses. Then once they have the scaffolding, we start to infuse the AI tool applications in.
A good example is Campus’s two stages of English. Our first course, English Comprehension, happens in the first semester, and it’s focused on making sure we know where students are in reading, writing and grammar. Then we move into Advanced Composition, at which point students should know how to effectively utilize these technologies. In fact, that course is all about leveraging generative AI tools. We make sure students have a rock-solid foundation without these technologies, but we then layer it into the curriculum to elevat the work even more.
Focus assessments on evaluating the inputs and process rather than just the final output. And maybe even consider raising the bar on assessment.
Asim: One example of a process tweak we’ve put into practice is shifting to grading inputs and process over just the final product. Let’s say you have an essay assignment—what if we asked students how they chose their topic or how they went about identifying their sources. If we can focus on the elements that go into building the final product, then it not only builds better scaffolding for the assignment but it also leads to a more transparent assignment design. Also, it puts instructors in a position to identify aspects where it really makes sense to leverage AI. That can help students develop the skills to identify what they should do on their own versus where it makes sense to use emerging technologies.
Kaari: The standards don’t stay the same when students are allowed to use these tools. They skyrocket. That ten-page research paper with ten sources should now be a 20-page research paper with 20 amazing sources and argued from multiple points of view. So, the bar is raised in order to make sure that the work that goes into a final product is the same with and even more elevated outcome.
Separate curriculum into perishable and nonperishable components. Focus on updating the perishable parts more likely to change quickly.
Kaari: There’s a balancing act between the nonperishable and perishable content upon which we build our courses. We need to identify the fundamental outcomes we know students need to learn, then figure out the pathways to achieving those outcomes, which will evolve and change very rapidly. I see our role as ensuring there is a strong core, and that’s about enabling learners to think critically and having the confidence to learn these new tools as they evolve.
Be Open to AI’s role outside the classroom and core curriculum.
Jessica: In speaking to employers, we actually see less use cases for directly writing code but more AI tools adoption for helping support the day to day—the extra practices that go into being a good product builder or leader in software. So, we’re also taking that approach. We’re asking our instructors to tinker with AI tools to help our students get faster and more effective at those ancillary parts of the job.
Asim: We also see AI’s benefit beyond the classroom. At Auburn, we have tutoring in place to help at-risk students, but there are just never going to be enough resources—the number of tutors and their availability—to serve every single learner and help them reach their potential. So, what if AI could help us reach all learners and not just at-risk students?
While AI tools are developing rapidly, the key is to focus on the fundamentals that do not change: critical thinking and foundational skills related to each discipline. Educators and program teams can then experiment with AI tools in small ways to build familiarity and determine where they can enhance learning without replacing essential skills development. With these lessons in mind, educators can navigate the evolving AI landscape while continuing to develop the most important skills in their students.
Tools and Resources mentioned by the panellists
- Auburn University’s full online, self-paced AI Course for Faculty
- GitHub Copilot—AI developer tool
- The Sentient Syllabus Project
This is the part of a series of events on Harnessing AI in Adult Education. If you enjoyed reading this recap, consider joining the Learner Success Guild, a community for online education professionals to learn and share best practices on supporting adult learners.