Critical AI Literacy and Flat Information

[mimmoore]

Several years ago, I developed a Quality Enhancement Plan (QEP) for the community college where I worked at the time. Our topic was information literacy. Fortunately for us, generative AI was not yet widely known or accessible to students (or to the faculty QEP team). We used the Association of College and Research Library’s Framework for Information Literacy for Higher Education as one of our guiding texts.

I had already seen how the internet had made information “flat” for many of my students: they did not seem to understand that information made its way to an online source via multiple routes, with varying degrees of transparency and accuracy. In short, for some students, all things lived online in basically the same way—in a flat, two-dimensional information landscape, lacking contour, context, or layers.   

The Information Literacy Framework, in contrast, proclaimed that “information creation is a process;” developing information literacy entails strategic evaluation—of why, by whom, for whom, and how—a text, video, purported fact, or image was created and shared. 

Fast forward to the present. The advent of widely available generative AI (by which I refer to large-language models that include ChatGPT and Google’s Gemini, among many others) has exacerbated the conception of “flat information” in ways that I could not have envisioned.  In fact, a colleague recently commented that in some fields, it really doesn’t matter if an artifact is human or AI-generated; what matters is what students can do with it.

I was stunned by that comment.  Granted, this colleague was not speaking of “information” per se; the focus was on products such as reports, posters, datasets, tables, etc. But it raised for me a central question: does it matter whether a product or a piece of information is generated by a machine?  If so, when does that provenance matter? And do I have the right to know when text (and the information in it) is the output of an algorithm or machine learning?

The proliferation of fake images and misinformation following Hurricane Helene and Hurricane Milton has been mind-boggling.  Do my students know how those images and “reports” came to be or how to verify their accuracy?  (Do I?) When students search online and get a coherent answer from Google’s Gemini, do they understand that it was produced based on statistical patterns of language data, not on a search of facts? Do they recognize the disclaimer that follows Gemini’s output?  Do they know they can look at the blue boxes on the right and find Gemini’s sources (which still need to be understood in the context of who, why, how, and for what audience)? Do they distinguish between tools and information sources? 

This fall, I am trying to blend a writing-about-writing syllabus with a writing-about-generative-AI syllabus.  I want my students to see that just as their writing is the result of a process (of thinking, collaborating, drafting, using tools, fact-checking, revising, editing, and other things), so also AI came to us via a process, and it has added new layers to the processes of creating and disseminating information. Asking questions about these processes—and recognizing them as processes, not just a landscape of flat products—seems to be a reasonable response to technological changes that I cannot keep up with.