Start Small
Start Now
Getting started with an AI literacy curriculum seems like a daunting task.
Find one small thing you can do this semester.
It seems like a daunting task: learn AI, figure out what to teach, figure out how to teach, overhaul your syllabus, heck, develop an entirely new pedagogical approach!
Integrating AI literacy in one's curriculum can become a struggle to survive in the quicksand. At the same time, the cost of inaction in today's rapidly changing technology environment is too great. So this is what I suggest: find one small thing you can do toward AI literacy this semester.
The trick is to start small but keep the option for future expansion open. To do this you will need a conceptual roadmap to guide you through your own AI literacy development and gradual curricular integration. I developed the "AI To Lean" framework with this kind of curricular building process in mind. You can start with one activity right now. It won't change student practices overnight, but you can at least expect a meaningful discussion about using AI productively and responsibly in the academic context.
For example, in some recent workshops, I shared a simple class activity called "What is a large language model?". It starts by getting students in the classroom to use the same prompt to elicit AI response, then discussing what we found. It will take 20-30 minutes of class time, and yet, I found it to be a far more effective way of teaching the very basic mechanism of how AI tools generate its responses, and how human judgment plays a critical role in that process. It can open up a whole new perspective on generative AI and what it takes to engage it productively and responsibly, and unlike more passive forms of information transmission, like reading an article or watching a video about the current transformer architecture, students remember what they got out of the hands-on exercise.
Learning to Teach: An Anthropologist's Journey
Sawa Kurotani (AI To Learn Institute) recalls her journey of building her own AI literacy curriculum from the scratch with no prior technical background.
"I'm a humanistic anthropologist, trained in qualitative research. I taught interdisciplinary courses in social sciences and humanities, where reading, writing and talking are the primary academic activities. I considered myself decent at using common technology tools in her daily life - check emails, write lecture notes, watch movies, stay in touch with friends via social media - but have no training in IT, computer science, programming, etc. I was clueless about what actually happens inside my desktop computer or Android phone when I type words on my keyboard..."
"I went through a period of panic when I decided in early 2023 to look into what the hype around generative AI was all about, primarily to figure out how this new technology would affect my students' learning and how I would need to adjust my teaching strategies. Beyond the most generic marketing spiel on OpenAI's public-facing website and some pretty pictures made with their image generation model, very little I read or saw made sense."
"It got worse when I got to OAI's Discord channel. I looked through the general conversation on ChatGPT, and I could understand maybe one out of 10 messages. And the posts kept coming in one after another as she I watched. I went back the next day and the same thing. I went back again and this time, I scrolled down and found the chanels dedicated to AI-generated images, which unexpectedly became my entry point into the AI universe (more in DALL-E Legacy if you are curious what happened)."
Y"ou may be thinking, 'What does AI image generation have to do with teaching AI literacy?' Not so fast! I'm just trying to say, before people like me (and probably you, too, if you are still reading my story) to get started with AI, we need to find an entry point where we can find something to hang on to for the rest of the ride (which, I promise you, will be turbulent). I learned the quirky ways of generative AI and got started with essential skills, like prompt engineering, which I was able to transfer to other areas like writing, research, and teaching."
After having become an active member of multiple AI communities and developed her own human-centered AI literacy framework, Sawa's final words are:
"Those of us from non-technology background has the advantage of having started this journey recently, getting lost, and feeling confused. All these experiences become our biggest asset when the time comes to help students, many of whom will struggle, as we have, to begin their own journey..."
"Another distinct advantage of the technologically disinclined is the fact that we know our stuff - our content knowledge and content-specific pedagogical skills. That is, I know how anthropologists write; I also know how to teach students to think like an anthropologist and write like an anthropologist. Technology piece is an assistive element in this pedagogical process, and neither I nor my students need to know everything about the technology tool. Rather, we need to learn "just enough" about it to be a savvy user in the specific context of use."
Curricular Integration: Incremental Approach
Getting started is the hardest part of integrating AI literacy in one's courses. When I decided to do this in 2023, there were few usable curricular resources available. I ended up developing one from the ground up, iterating and improving it over two academic years.
In the first academic year of implementing AI literacy curriculum, I took a modular approach: three 30-minute AI-literacy sessions in every class I taught that covered a very basic introduction to responsible AI use. In the second year, I began to experiment with a more integrated approach. Students still went through the foundational modules, but they were also asked to use AI in the stages of some graded assignments, for example, using AI feedback in an essay revision process or conducting AI-assisted data analysis. AI literacy, in other words, becomes the foundation upon which students can build their context-specific competencies to leverage AI and accomplish real-life tasks more efficiently and accurately. By the fourth semester of integration, I was able to teach a fully AI-integrated course. In the third year of implementation, all my courses are fully AI-integrated with one class session dedicated to AI literacy introduction.
If you are thinking about incorporating AI literacy training into your curriculum, I strongly recommend starting with the modular approach first. In most cases, we can accommodate three 30-minute sessions with a modest amount of syllabus/course plan revision. Try it for a semester or an academic year, and move gradually to a more integrated approach, keeping the amount of curricular work manageable.
If you are not ready to go there yet, consider making just one small move this semester: brainstorm where in your existing syllabus/course plan you can potentially introduce one guided activity in which students will encounter generative AI tool in a structured, thoughtful way. If you are interested, click the button below for more quick-start resources.
Managing Information Overload
It is important for all of us to stay updated on how AI technology is evolving. But the amount of information about AI development and its future implications are voluminous and complex, and expert opinions are diverse and often conflicting. "Start small, start now" applies to this aspect of teaching and learning with AI. Here are the things I keep in mind so that I don't get overwhelmed and keep my head (sort of) above water:
-
Forget the idea that I'm going to ever "keep up." There is just so much stuff out there and there's no way anyone can know about everything out there. Besides, AI technology keeps evolving rapidly, and a fact or a tip that was valid last week might become totally obsolite overnight. No one can truly "keep up" with all that.
-
Prioritize, prioritize, prioritize. Choose one or two very specific aspects of the AI universe and ignore everything else. My two priorities are AI literacy pedagogy and image/video generation. I try to keep an eye on any new information that are relevant to these priority areas and relentlessly skip, delete, and ignore everything else that pops up on Google search or in my email inboxes. Trust me, you are still going to have a plenty to read and watch, and because everything is interconnected, you'll still be learning about other AI-related topics. But at least, you know why you are reading them - because they are somehow relevant to what you are trying to do in your priority areas.
-
Find a few go-to sources where you are most likely to to find news and information that are most helpful to you. Be totally selfish and utilitarian: you keep whatever work best for you or seems interesting to you and set aside the rest. It will take time to find the right ones for you, but it's necessary investment to keep your sanity in a long run. My current go-to's are the Discord servers of Google and OpenAI for latest tech development in image/video gen, a few subscribed channels on Substack and YouTube that focus on AI literacy and technology breakdown.
Curriculum-Building Workshops
Intro to AI Literacy Faculty Workshop Slide Deck (PDF) - intended for faculty and teachers who are considering AI literacy integration, this slide deck outlines the AI To Learn framework, a guided inquiry exercise, and how to start implementing AI literacy curriculum.
Cultivating Judgement Faculty Workshop Slide Deck (PDF) - intended for faculty and teachers who are considering AI literacy integration, this slide deck focuses on the concept of "workflow" to consider how to integrate AI assistance responsibly and productively in an assignment process.