Writing, AI

 I haven't been exploring AI as much as I feel I should be. 

Because my job is about teaching writing, the set of LLM-based platforms that have become mainstream since Nov. 2022 (ChatGPT, Claude, Gemini, Co-pilot, ...) will become a Very Big Thing in my professional life. I do fully expect that they will entirely transform teaching writing as it currently exists--from pre-K on.

Early on (Spring 2023) I helped facilitate discussions and a speaker series, hoping to provide some guidance to faculty as they adapted to the availability of these AI applications. But, aside from trying to develop an understanding of generative AI, I haven't spent much real time trying to figure out how to apply them to my own work tasks. Or, rather: when I have tried, I've been pretty quickly discouraged at how clumsy ChatGPT has been at responding to my requests. 

By its nature, generative AI manages a mostly conservative response to queries: It generates text based on reasonable new associations based on existing word relationships: things already "networked together" are going to be re-packaged but, typically (in my experience), not transformed with nuance. 

Something I've been telling students is that this technology favors those with expertise. Simply put: You have to know something in order to tell if ChatGPT is producing bullshit or not. 

I understand that some of the powerful potential of generative AI is tied to prompting: the better a user's prompt, the more nuanced the tool becomes at assembling and weighting associations as it produces new text--and hypothetically, then, the text it produces should then be more appropriate to the situation. 

On the bright side, maybe: Rhetorical expertise is the thing that should help people write more effective prompts AND evaluate the rhetorical effectiveness of output from platforms like ChatGPT. This looks different than old versions of teaching in some substantial ways, but it may also allow us to raise the bar, as well, if it means we can shift some instructional time and student practice from generating basic blocks of text and on towards evaluating audiences and products. 

Because I value writing-as-process, I think writing teachers need to pay close attention to what is gained and lost in shifting towards new ways of asking students to produce rhetorically effective text. Students are still "producers," but in ways that may have some pretty big impacts on identity, ownership, and engagement. 


Comments

Post a Comment

Popular posts from this blog

What "Ordinary Men" teaches about conditioning for mass murder

Ordinary Men: Reserve Police Battalion 101 and the Final Solution in Poland is a book about a Hamburg-based police force's involvement in the mass killing of Jews during WWII. It is, in part, a historian's effort to understand how relatively "ordinary men" (conscripted into the war effort relatively late in the overall timeline) became willing participants in execution. "Willing" is perhaps not the right word, and in fact many did not participate on account of deep ideological values -- but the interviews on which the book is based do point to the variety of justifications and manipulations that made it possible for these ordinary men to become killers. Here's what I've learned so far: Collapse the distance between "police" and "military." Make killing legal. Create conditions of economic uncertainty. Create promotion ladders (or, alternately, threats against career promotion for non-participants) Make opportunities for "con...
Read more

What I think I know about writing (pt I)

"there are no general recipes or designs that can be applied to all contexts, but certain principles and experiences that must be adjusted to the predominant theoretical and pedagogical traditions, to the profiles of instructors and students, and to the institutional resources and to the curricular spaces available." -- Federico Navarro, "Think Globally, Act Locally: How to Design an Academic Writing Course for Students Entering University," p. 112 (https://doi.org/10.37514/INT-B.2025.2739.2.04) as Rose and Martin (2012) point out, if students are not taught to read and write independently, educational institutions will only reproduce the underlying social and cultural inequalities of their students. (Navarro, p. 119) Writing encourages a slowing of thinking. That slowing down is good when it allows for inspection and growth. To the extent that a process approach encourages inspection, it is good for writers' development. A process approach that produces change ...
Read more

FYW Curriculum Idea 1: Federal Program Impacts

 A motivating question: How can we get writers to move beyond ideology to explore, understand, and make decisions based on evidence? EXIGENCIES/ASSUMPTIONS Immediately (Spring 2025), DOGE is massively cutting federal funding for a variety of agencies and programs that have existed for decades or more.  Broadly speaking, citizens (of all ages) are getting "further from the facts," in terms of forming attitudes about civic behavior. There is a deepening of inductive (top-down) thinking about the role of government, where sweeping political ideologies over-drive people's orientation to specific events. (The alternate I prefer in this CI is deductive exploration of information, a kind of "working up" from evidence, ideally to encourage subtle interrogation of top-down ideologies they may have already acquired.) GENERAL SKETCH A note: I think this CI is probably "unit scale" rather than "semester scale"--it would probably need to be combined with...
Read more