Source: Photo by Héctor J. Rivas on Unsplash
As part of research for an upcoming publication (pending the time to dig deeply), I had the great pleasure of spending a couple of hours of my day with Bozkurt et al. (2024)––and what a group of et al. on The Manifesto for Teaching and Learning in a Time of Generative AI: A Critical Collective Stance to Better Navigate the Future. As is my habit, as I read the printed copy and added a lot of notes to the margins. A lot. Primarily, my notes reminded me about the relevance of sections and phrases to my research (a project on graduate learner use of generative AI (GenAI) for digital resource design), but then I started writing some notes that were simply critical thinking about what was being shared and the system issues in post-secondary education that I experience and observe regularly in my work related to GenAI. My current work is consulting for strategic planning, curriculum design, online teaching and learning, and the use of technology in education. So, all of that is well related to the Manifesto and I am so appreciative of the hard work of the authors who brought forward critical thinking related to the use of generative artificial intelligence (GenAI) in education. I have read a lot of articles in the past year and this one was the best. Comprehensive, balanced, and appropriately critical.
My initial advice related to this reading is…read it for yourself. I found it extraordinarily useful and informative and it covered most of the current critical thinking ground an educator might want to cover about the emergence and rise of GenAI over the past two years. One of the first items covered in the manifesto was the concept of metaphors where AI is sometimes being called a “co-pilot,” or “external brain,” and more negatively perhaps a “demon,” a “bullshit generator,” or, a “weapon of mass destruction” (Bozkurt et al., 2024, p. 488). My margin note, what if we called it the chains for capitalist vendors to enslave institutions, faculty, and students? I believe it’s a fair metaphor for a critical thinking conversation about GenAI. The authors also asked “How can we ensure that the design and development of GenAI align with our values, fostering care, equity, and inclusivity in education and beyond?” (p. 489). My note, in what ways do we have control over elements like that? What I was really asking there was, have I been invited to ensure anything at all about the design and development of GenAI? Not yet, as far as I know. Sure, I could make my own GPTs using some of the paid tools, but do I really know what I’m doing? Not yet.
The authors provided a good frame of where we are with AI (my note), as follows: “The introduction of GenAI technologies, which utilize large language models (LLMs) and leverage natural language processing (NLP), most notably exemplified by ChatGPT released by OpenAI on November 30, 2022, has ushered in a new era characterized by a blend of excitement, hype, hope, and speculation, particularly within the context of education” (p. 489). In an early section of the article, the authors identified (through a rigorous process of collaboration) a list of 15 AI Positive themes. Some of the themes and related critical insights inspired more margin comments. “Efficiency should serve to enhance education,
not to diminish the role of personal connection, mentorship, and critical thinking” (p. 491) to which I responded, do not lose human connections to efficiency. For this one, “Customized instruction may unintentionally reinforce biases if based on skewed datasets, limiting diverse perspectives, particularly in culturally sensitive areas” (p. 492), I wanted to confirm for myself may introduce or reinforce bias, don’t skip deeper learning, and what about Indigenous knowledge and ways of knowing, are they accounted for in personal learning and personal tutoring in current GenAI tools? Any other ways of knowing that are missing?
In the context of the thread of GenAI learning as important for future workplaces, this advice was provided, “The ability to critically assess GenAI-generated outputs will likely be key in maintaining human oversight and preventing over-reliance on GenAI-driven automation (p. 493). I asked, in what ways might we be teaching learners to critically assess GenAI-generated outputs if we’re not allowing them to use GenAI at all, or only in limited ways that we decide (or that our institutions decide via policy)? In this next example, I feel like the rubber really met the road, “Its rapid development has triggered much-needed reflection on what value teachers bring and what kind of learning outcomes are most meaningful. This re-evaluation is not just about improving efficiency but about re-focusing the purpose of education” (p. 493). Here I asked, what value do teachers bring to learning right now? Can they articulate it? Can institutions articulate it? How might faculty help re-focus the purpose of education in partnership with learners? How indeed?
Circling back to the value of educators, this one grabbed my attention as well, “…while GenAI could aid lesson planning and content creation, the quality and relevance of AI-generated materials require careful scrutiny to avoid generic or biased content that fails to meet diverse student needs and fails to reflect teachers’ unique positionalities and intersectional identities” (p. 494) and here I noted, AI may be too generic or too biased. This makes it a poor reflection of educator excellence. What elements of excellence can educators bring to the table when thinking critically with learners about GenAI? In the same paragraph, the authors went on to confirm that GenAI has to promote deep learning and critical thinking (a repeating theme in the article) – agreed. Here’s one that really knocked me over related to systemic problems, “To foster genuine autonomy, students need to be challenged beyond what GenAI can provide, balancing support with opportunities for intellectual growth” (p. 494). My note was this (with apologies as I don’t mean to offend), are faculty effectively prepared to go beyond what GenAI can provide? Do we have the rigorous training and subject matter expertise beyond a generic level? If yes, what does “going beyond” look like? If no, how can we begin to better ensure it? I am not certain about the rest of my global colleagues in academia, but I sometimes teach courses where I am not a full expert. I have seen a few instances where GenAI has outsmarted me (in its own charming and ill-informed ways). How can I guide learners appropriately related to GenAI accuracy if I’m not a genuine expert?
In the AI Positive section, I had just one more key thought, related to this quote, “Cognitive strain is part of the learning process – essential for developing resilience, critical thinking, and the ability to adapt to new situations. If GenAI shields learners from this necessary struggle, we risk producing graduates who are adept at using tools but lack the cognitive depth to navigate novel challenges independently” (p. 498). My system question here was, if cognitive strain is useful for learning, why do we punish learners with poor grades when they have engaged in cognitive strain (stretched themselves and their skills) and not achieved our institutional ideals for rubric-defined perfection? It seems to me that if GenAI can help me as a learner to easily align my output with institutional ideals, I’m more likely to take that easy path.
Moving on to the AI Negatives portion of the article, there was a section on Ethical Considerations. I have recently talked on LinkedIn about how challenged I am by the theft of knowledge from the whole Internet by GenAI organizations (and here I am looking directly at OpenAI as the principal offender) and the nearly complete silence (and I would posit complicity) in the world about copyright law related to this egregious and continuing paradigm. A key quote is, “…repurposing creators’ work without consent threatens the sustainability of creative industries and erodes trust in AI. Beyond the creative sector, GenAI’s ability to process vast amounts of personal data without users’ knowledge raises serious privacy concerns. Individuals often lack control over how their data is used, which highlights the need for transparency and stronger regulations” (p. 499). What I had to say about that is, we currently have robust policies, privacy regulations, and laws about copyright. Why in the world aren’t they being enforced? Where is Disney’s historic outrage? If these copyright and privacy policies, regulations, and laws mean nothing, what other standards and laws mean nothing? Can I stop paying my taxes now? Can I speed around my town with impunity? Oh, dang, I left my entitled white man ID in my other wallet.
As you might guess, related to Commercialization and Concentration of Power as a negative risk of GenAI paradigms, I had a few notes. One quote was, “Beyond job loss, there’s a risk of devaluing human expertise” (p. 502) I asked, what is it that humans do and GenAI cannot do, and why does it matter? And also, this risk has been raised for generations related to computers and technology, why does it feel new right now? General responses in this section included, why are big tech investors sinking billions of dollars into GenAI? Of course corporations are going to be prioritizing profit over educational outcomes, what are the motivations of Microsoft, and most LMS, design, and textbook publishers adopting and adding AI “helpers” to their products? What is the bandwagon everyone is jumping on and why? FOMO or some other more nefarious motivation?
Toward the end of the AI Negatives section, I underlined this one, “Resistance to change is another challenge; educators may feel that AI threatens their autonomy and teaching methods, while students may view GenAI tools as impersonal and diminishing their sense of agency” (p. 506), and my thought was, if this article is any indication (which it is), such resistance is justified and should be encouraged. I dislike the positioning of critical thinkers as resisters or laggards––sometimes what they are resisting is dangerous and useless change for the sake of change in education. In the conclusions, I had just a couple of ideas, “First, GenAI is not just a tool; it is often treated as an agent with capabilities to communicate, interact, and create content on demand, despite differences with human-to-human processes” (p. 506), here I had a thought, I have framed GenAI as just another tool a few times recently and I want to start walking that back. The authors are right, it’s not just a tool because of the way it is being framed by vendors and hype experts (and even many in our own institutions) as an interactive intelligence. Folks can see part of the intention of these “tools” is to take away jobs, in particular well-paid educator and creative jobs. This nonsense about freeing up educator time for more important tasks is just that, nonsense. Jobs will be lost. I’ll leave you here with my agreement with the authors’ positive question, “Do we follow the white rabbit blindly into a world dominated by GenAI, or do we approach it with wisdom, caution, and purpose? (p. 508), yes, we do approach it with wisdom, caution, and purpose and we are all better off with authors like the collective that created this super useful document. What are your thoughts and margin notes?
