Allison Stanger is objectively smarter and more successful than I. She's a distinguished professor who has published books and articles in major outlets, testified before Congress, and just generally done a lot of impressive stuff.
And yet, I find her recent article for The Chronicle of Higher Education just incomprehensibly terrible.
"
Who Cares Whether Academics Use AI to Write?" is yet another one of those "I like AI just fine cause it's such a big help" articles that leaves me scratching my metaphorical head.
Stanger's basic point is that in the world of scholarly writing, an argument is "meant for readers to understand, contest, and/or deploy." And if the argument is solid and the point is worthwhile, then why should anyone care where it came from? So, scholarship without scholars? Stanger argues that writing can be judged by considering its ultimate purpose, but her implication is that the ultimate purpose of some writing does not require a human being on one end of the exchange. Which begs the question-- do we need a human on either end? If the AI extrudes an argument that logically makes a useful point, would it be enough if that scholarly insight was picked up by another AI to use in extruding the next piece of smartitude? Does the reader that understands, contests, and/or deploys the scholarly argument have to be a human reader? Can it just be another AI scanner? And if the argument is that it needs to be a human reader, why does it not have to be a human writer? If a gaggle of AIs crank out a series of scholarly arguments and no human ever sees or hear them, do they matter?
Stanger really wants to make an argument for collaboration. Exhibit A kicks off the article with the story of an article shed wrote for a professional journal which they wanted to have re-edited so it would include more of "her voice" and offered to do the work themselves, and she is correct to point out the bizarreness of having "strangers rewrite my sentences so that they would sound more like me."
But they weren't entirely her sentences. The article went through ten drafts that she worked on with Claude Fable 5.1 and "other large language models, as I now work on nearly everything." But she reviewed every line and rewrote most of them. "By draft 10," she writes, "I could not have said where the machine’s contribution stopped and mine began." Really? I find that assertion incomprehensible. And if you rewrote most of the lines, why have AI write them to begin with? But Stanger asserts that the editor liked the argument, evidence and structure just fine, without making a case for who or what provided those features. Maybe she doesn't remember, but how the hell do you write a piece if you haven't already worked all of those things out in your own head?
I don't get these sorts of pieces. What is the point of adding a janky plagiarism machine to your writing process. In the larger world, the pitch often provides a simple answer-- the work. "I wanted to write a book and it was just too hard but now ChatGPT can write my book for me and I can skip all the hard part." But for academics like Stanger, the answer is often hidden. And true to form, one thing that Stanger does not provide is some statement of "Now that I use AI, I am able to [fill in the blank] more effectively than I could working on my own."
I always want to ask these folks the same question-- what can you do now with the AI that you could never do before on your own?
Exhibit B is this very article, which she says she produced by the same process. Ten drafts? With "extended collaboration" with AI? The piece is just under 1300 words and zero academic features like cited sources. And it includes paragraphs like this one:
How we educate our students in this brave new world is a question unto itself. The student essay was designed to provide evidence of learning, which it no longer does; a novice writer cannot responsibly direct a tool whose output they cannot yet judge. Teaching therefore requires careful scaffolding, which can be constructed in myriad ways, including a ban on AI use.
I mean, that's not terrible, but it's not ten-drafts-with-a-gifted-collaborator either.
There's one more point to note in this article-- the really bad news for students (I may have buried the lead here). Stanger says she now teaches with AI (whatever that means) "and that has meant replacing my old assignments with harder ones." After noting her devotion to some human activities like close reading, rigorous citation, and Socratic method, she drops this sentence:
Because AI speeds up the routine work, I can assign more of it, in more varied forms: research, outlines of arguments, revisions of their own prose, oral presentations, and slide decks.
In other words, she has designed a course so that an unaided human, a student who has chosen to avoid AI "assistance," faces an almost unmanageable workload. The answer to "What can I do as a teacher that I couldn't do before AI" appears to be "Bury them under a literally inhuman workload." This skips right past study of AI and digging into the implications of using it and jumps ahead to "use it or suffer the consequences."
Stanger doesn't appear to have completely lost the plot. "For all forms of writing, readers are owed a human being who stands behind the work," she writes. "A byline has always signaled ethical and legal responsibility." But not unassisted composition.
It sure looks as if Stanger has wrestled with some of the above issues elsewhere- her next book is Who Elected Big Tech, and she is a founding member of the Digital Humanism Initiative. But I find articles like the one covered here just kind of sad and discouraging. Writing is, to me, a supremely human, almost sacred action, and I am no more convinced that genAI can enhance the experience than I am convinced that getting a robot to kiss my wife for me would improve our marriage. I am, however, convinced that if people keep telling themselves that they need a bot to help them write, they can make it so. And that makes me sad.
No comments:
Post a Comment