I was recently asked to provide a short back-of-cover blurb for Alym Amlani and Paul Davis’ new book: ChatGPT Ate My Homework: What Educators Need to Know about Gen AI.
The endorsement: Generative artificial intelligence has left many instructors feeling powerless, especially over their assessments. This insightful work models how to respond to generative artificial intelligence so that instructors can harness the learning opportunities these tools provide and ensure students don’t sidestep the learning process.
I was grateful for the invitation to review this eminently readable book. It models good behavior using genAI in the writing and teaching process, and it captures the real concerns many educators have. Educational resistance is not just Luddite techno-skepticism.
After reading the book, here are three things I am left wondering about: :
Is it time to ditch Bloom’s taxonomy once and for all? Roland Case (2013) called Bloom’s taxonomy “one of the most destructive theories in education,” and I agree with Case’s analysis because there is not (as far as I am aware) any empirical evidence supporting Bloom’s taxonomy. Bloom’s was initially intended as a theory of assessment, not a theory of teaching, and while I adhere to the motto that “all models are broken, but some are useful,” I am not sure how useful Bloom’s is anymore. In the wake of genAI, should Bloom’s be thrown in the graveyard of bad ideas next to the myth of learning styles?
Bloom’s has locked in as a theory of teaching because there is a truthiness to it; most people assume there are some lower-order thinking skills (memorizing and recall) and higher-order thinking skills (synthesis and evaluation). GenAI can do much lower-end “thinking” (and some pretty good higher end “thinking” such as comparing and summarizing), but how can educators start students at the higher end of the scale? If students don’t understand the basic conceptual vocabulary often taught at the “lower levels” of the hierarchy, how can learning ladder or be scaffolded? How can students evaluate concepts they don’t understand? As Case points out, “assessing students’ ability to complete the ‘higher oder’ tasks does not logically imply that students have mastered the ‘lower order’ task.” Amlani and Davis point out that teachers are concerned students will bypass a ladder, but the Bloom’s ladder does not, in fact, exist. These domains are integrated, and I quite agree with their assertion that Bloom’s looks dated. Still, they provide an expanded triangular hierarchy that continues to represent these competencies or skills as discrete blocks. This isn’t the breakthrough model we need. I don’t know what is, but I see concentric circles…Critically evaluating the results from genAI will require mastery of all these Bloom’s domains at the same time.
Generative AI is not neutral – Amlani and David describe genAI as a series of neutral tools (p. 68). This represents an instrumentalist view of a technology that has values built in, which they also recognize. There is an irreconcilable contradiction here. Instrumentalism and Interactionism and are not incompatible, per se, but instrumentalism is seriously deficient. Interactionism would help focus the educational purpose of using genAI as the creation of meaning (not just jumping through a set of assessment hoops). Data is the lifeblood of AI, and ingesting data is one of its values (hence the proliferation of free versions). Profit is another value. Reciprocity is not. Some AI designers have as their stated goal putting people out of work (not just predicting that it is likely to happen, but that they WANT it to happen). What are the values that reside at the foundation of genAI? What are the values that reside at the foundation of the educational enterprise? How are they at odds? How are they compatible? How does this inform how we approach genAI from an educational perspective? Yes, genAI is here to stay and we should use it appropriately for educational purposes, but whatever appropriately means, it means putting the values, not the tool, first.
Changing the game and changing the structure - Our brains are wired to seek out the best tools to earn the rewards we want. So change the nature of rewards. The holy grail that many educators have expressed to me over the past decade is that they wish they could move their students from a focus on grades to a focus on learning. Some of this work can and should happen at the micro-level (meaning within courses). I stand by the quote they use on page 154; students are largely asked to do things they don’t care about. But Amlani and Davis also include lots of good examples of occasions when students were asked to provide meaningful personal input in the learning process at no cost (couldn’t get a failing grade), and they still took the easy way out. If we did away with grades tomorrow (as Eyler suggests in Failing our Future), would that be the structural masterstroke?
In short, genAI cannot be introduced into this ecosystem under extreme stress without some thought for how the system needs to be restructured in the wake of genAI. Massification and the significant growth of part-time faculty forces instructors to use certain types of assessment for the appearance of objectivity and time management. Rewarding creativity and originality changes the nature of the game, but how much can be done within the current structure? I agree it should happen and know that it can happen (for example, I met an Information Technology instructor who has “Daring” as a major part of his rubric)., but can the scale of change we need to adapt to the impacts of genAI happen within the current structure?
GenAI use will be normalized, for sure, but if we are not in control of the values-based adaptations that need to take place to account for this new technology, higher education will continue to be molded to meet the needs of artificial intelligence, not the human intelligence we are supposed to cultivate.
These are three things I wonder about as I consider the best course for academic leadership to set in the wake of reading this worthwhile and helpful book.