Ironically, if anything, I find myself taking grading and reading student work even more seriously in the age of AI. I've also tried to vary my prompts and assignments more. I agree with you - if you don't want to engage with student writing, you shouldn't be assigning student writing.
100%. I had this sensation pre-AI when I changed my assignments so students were doing something we both thought was interesting. My time spent grading didn't change all that much, but the time spent was much more enjoyable because of the nature of the experience. If you know what you're doing matters, you put more into it.
I agree that teachers should not use LMs to “grade” student work for a variety of reasons. Teachers should give abundant feedback separate from evaluation with stakes attached. I’m less inclined to argue that students should not use LMs for feedback. There is fertile ground on this front I believe for students to learn about the nature of feedback. But I have lots of problems with the idea that a core function of a writing teacher is to “grade” student writing. Grading and feedback are separate actions. Grading is an institutional requirement, not a pedagogical requirement. Grades are as big of a problem for writing instruction as the use of AI to do the grading is. Statements like “teachers who use AI to grade writing should stop or quit their jobs” are hyperbolic and paper over the underlying problem, which is relying on institutional power (grades) as an essential teaching tool. I have no secret sauce to get rid of grades except to do it as little as possible, but we have decades of peer-reviewed research supporting the assertion that grades overpower intrinsic motivation (the most potent kind of motivation) and reshape how learners go about their work from getting better at the work to getting better grades, which are different motives and require different learning strategies.
Terry, as usual with me, you pop off without knowing my work. I am a longtime advocate for alternative/ungrading and not putting letter grades on student work. These notions are central to not just one, but two of my books on how we teach writing.
What I am responding to here is the practices of some teachers (that Marc Watkins documents) who have indeed turned to AI to grade student writing. These people should either stop - and explore alternatives to grading - or, if they don't think they can so, of if they sincerely believe AI grading is a good thing, they should quit.
I’m not “popping off” nor responding to you personally, John. Everything is not all about you. It’s about ideas. Grading, feedback, and recommending that teachers quit if they don’t follow John Warner’s preferences are ideas. We still have free speech.
It's hard to begrudge overworked teachers and college instructors using AI to grade formulaic homework, but when it comes to writing I 100% agree. Students need to write with an intentional reader in mind -- someone to whom they're trying to communicate something. I hear of students who say there's no point in writing because AI will do it for them, and also no point in reading because AI will tell them what they need to know, so we're looking at a world where AI writes for AI. Obviously that's absurd, AI can surely communicate with other AI more efficiently in binary code or something, so we're looking at the demise of writing and reading and, as so many people on Substack have said, all the thinking that that requires and encourages. (Depending on the day, I range from "the sky is falling" to "a lot of the sky is falling.")
Since the rise of LLMs, I have been assigning more in-class writing. Initially I did it because I cannot police students' LLM use outside the classroom, but then I discovered a very welcome side effect: I am much more motivated to give detailed feedback when I don't have to wonder whether the student actually wrote the thing I'm grading. And students don't feel pressure to fluff and pad their writing when they only have the class period to do it.
Now, I'm not a composition teacher, I'm in a STEM field. I use writing assignments to assess student understanding and to give them practice communicating their understanding of technical matters, which they don't get enough of. I don't know how good of an idea it is switching to in class writing in classes that are centered around writing itself. I think it's made my classes better, and I wish I'd done it before LLMs forced my hand.
Ironically, if anything, I find myself taking grading and reading student work even more seriously in the age of AI. I've also tried to vary my prompts and assignments more. I agree with you - if you don't want to engage with student writing, you shouldn't be assigning student writing.
100%. I had this sensation pre-AI when I changed my assignments so students were doing something we both thought was interesting. My time spent grading didn't change all that much, but the time spent was much more enjoyable because of the nature of the experience. If you know what you're doing matters, you put more into it.
I agree that teachers should not use LMs to “grade” student work for a variety of reasons. Teachers should give abundant feedback separate from evaluation with stakes attached. I’m less inclined to argue that students should not use LMs for feedback. There is fertile ground on this front I believe for students to learn about the nature of feedback. But I have lots of problems with the idea that a core function of a writing teacher is to “grade” student writing. Grading and feedback are separate actions. Grading is an institutional requirement, not a pedagogical requirement. Grades are as big of a problem for writing instruction as the use of AI to do the grading is. Statements like “teachers who use AI to grade writing should stop or quit their jobs” are hyperbolic and paper over the underlying problem, which is relying on institutional power (grades) as an essential teaching tool. I have no secret sauce to get rid of grades except to do it as little as possible, but we have decades of peer-reviewed research supporting the assertion that grades overpower intrinsic motivation (the most potent kind of motivation) and reshape how learners go about their work from getting better at the work to getting better grades, which are different motives and require different learning strategies.
Terry, as usual with me, you pop off without knowing my work. I am a longtime advocate for alternative/ungrading and not putting letter grades on student work. These notions are central to not just one, but two of my books on how we teach writing.
What I am responding to here is the practices of some teachers (that Marc Watkins documents) who have indeed turned to AI to grade student writing. These people should either stop - and explore alternatives to grading - or, if they don't think they can so, of if they sincerely believe AI grading is a good thing, they should quit.
I’m not “popping off” nor responding to you personally, John. Everything is not all about you. It’s about ideas. Grading, feedback, and recommending that teachers quit if they don’t follow John Warner’s preferences are ideas. We still have free speech.
Spot on! I'm also not about making my own job irrelevant. At the K12 level, AI has become the snake oil for everything.
It's hard to begrudge overworked teachers and college instructors using AI to grade formulaic homework, but when it comes to writing I 100% agree. Students need to write with an intentional reader in mind -- someone to whom they're trying to communicate something. I hear of students who say there's no point in writing because AI will do it for them, and also no point in reading because AI will tell them what they need to know, so we're looking at a world where AI writes for AI. Obviously that's absurd, AI can surely communicate with other AI more efficiently in binary code or something, so we're looking at the demise of writing and reading and, as so many people on Substack have said, all the thinking that that requires and encourages. (Depending on the day, I range from "the sky is falling" to "a lot of the sky is falling.")
Since the rise of LLMs, I have been assigning more in-class writing. Initially I did it because I cannot police students' LLM use outside the classroom, but then I discovered a very welcome side effect: I am much more motivated to give detailed feedback when I don't have to wonder whether the student actually wrote the thing I'm grading. And students don't feel pressure to fluff and pad their writing when they only have the class period to do it.
Now, I'm not a composition teacher, I'm in a STEM field. I use writing assignments to assess student understanding and to give them practice communicating their understanding of technical matters, which they don't get enough of. I don't know how good of an idea it is switching to in class writing in classes that are centered around writing itself. I think it's made my classes better, and I wish I'd done it before LLMs forced my hand.