We've lost the ability to project classwork into students' time at home, and mourn that loss -- and are reluctant to accept that and struggle to internalize it. We can keep doing what we've been doing, to the benefit of students still willing to do that work, or give up entirely and rely on in-class assessment during the time we have each week, or radically expand class time. That last one will meet tremendous resistance as it'll mean cloistering students for long stretches, but will only seem like an overreaction to anyone who hasn't come to terms with the magnitude of the problem. (Call me an alarmist but we alarmists are having our day.)
I don't think that's alarmist at all--it's just a description of what has happened. There are lots of things we're going to have to figure out as a result, but at the top of the list for me is figuring out what sorts of work--real, time-consuming, difficult, learning-producing work--we can give them to do outside of class. (Or, rather, meaningful work that we can give them and also have some reliable sense of whether they've done it and what they've learned from it.)
I think what makes this diagnosis feel alarmist is that we can obviously still give them the same meaningful work we've always given them (reading, writing, problem sets, etc), and *they can still do that work* and derive enormous benefits. The vast majority of students (or at least of my students) are still doing the work because their purpose in college is to gain a real education. It's what they want, so it's what they do. And as long as that's true, it feels alarmist to freak out.
But across higher ed, it increasingly feels like that norm is eroding. Many of my grad school friends, who teach at very different places, are downright despondent. The truth is that we've never been able to *make* anyone do the work, but there have been fairly reliable signals we can look for to see if they've done it and to assess what they've learned. The carrot-and-stick of grades was usually enough. It's not alarmist to point out that students no longer have to do the work we assign them to have something to turn in, and aside from gut-level suspicions or devoting serious class time to seeing what they know, it's much harder than ever to tell what they have and haven't done or learned.
Sorry for the long response, but I think you hit the nail on the head. What we need is new ways "to project classwork into students' time at home." The authentic work has always been "Read this. Write that. Work through this tricky problem." Not being able to ask them to do those things without a helpful chatbot popping up on the screen to ask if it can do it for them is heartbreaking. It might be great for "productivity," but it's a nightmare for traditional teaching and learning.
Our current class design assumes that students read and write at home. The humanities need to add on a “lab” component that gives us oversight on student work. It’s not just the writing but the reading that is being eroded (since students can upload pdf files).
Yes, I'm hearing more and more calls for something like this, and I can see some real advantages to having more time to do meaningful things together in class. (I also have some fears. I can't help but think about old-fashioned "study hall", the kind of labor arrangements we'd have to make to implement labs, and sending the signal that learning is something that can only be done with oversight vs in an independent, self-directed sort of way. None of these are insurmountable and the benefits might outweigh the costs. Until people try it we're all just guessing. But I still worry.)
I'm out of the undergrad teaching game for a few years but I've imagined revamped reading quizzes thare are so nickpicky that no AI-generated summary will suffice. Along the lines of, "Which of the following sentences was/was not in the reading?" But I'm reluctant to implement any change that will be hated by students without course evaluation forgiveness. Promised in advance, in writing.
I see the breakdown happening earlier. Building sentences and choosing words to imperfectly match meaning is fundamental to developing critical thinking skills. When my students realized how much they were missing — which they did when I taught “how to write a sentence” to undergrads, they mourned and celebrated at the same time. Why we are robbing them of this experience and exercise is truly beyond me.
Yes! Writing well and thinking critically (not to mention reading deeply) are all foundational and worth fighting for. I think that's why I feel so much grief for the way that AI is undercutting the traditional college essay. I'm in the lucky position that the vast majority of my students still would rather labor over their writing and learn something than use AI as a way to escape the work. But "vast majority" doesn't mean 100%, and right now I worry that what's protecting their learning is their motivation rather than anything special about what I'm asking them to do. Have you found any structures that help enforce the work aside from (in addition to?) giving them something to celebrate?
I'm interested in hearing what sort of alternate assignments educators are developing or pivoting to in light of this issue with AI. I've heard a lot about returning to blue books, but I feel like that will miss the students who don't test well -- which was why education had moved beyond them in the first place. What other options are there, old or new?
(For the record -- I currently teach outdoor school, and so I'm blessedly free of this debate at this point in my career. AI can't help you figure out how to work as a team in the middle of the woods ;) But I'm still paying attention, and interested in what this dilemma illuminates about the entire endeavor of teaching -- and learning.)
Thanks, KLB. I've also heard a lot about experiments with blue books, oral exams, presentations, and other more traditional testing methods. I know some people have had good success with that, but I share your equity concerns. Even before AI hit, I'd shifted during COVID to a combination of specifications ("specs") grading in big classes (regular reading responses, unit essays, etc) and in small classes stuck with traditional papers but added paper conferences (20-30 minutes per student with discussion and oral feedback from me). It's still mostly working--but the essay's opening anecdote is from one of the specs based classes. I'm actually wondering if what you're doing to get them to work as a team in the middle of the woods might have some lessons for us rather than the other way around! Anything you're seeing in places where all the screens are off that works really well?
In smaller seminar courses I've seen moderate success with forging personal connections with students, no- tech classrooms, oral exams, bluebooks, and an emphasis on creative material projects that require hands-on involvement (making an artifact, cooking, etc). But for large lectures & especially for online courses nobody seems to have good answers.
I think Anthropic, OpenAI et al should pay for a luxe summer camp for faculty invested in these questions.
In reality I think they owe Education so much more. They're disrupting an industry with which they are not in economic competition, as say Uber : Taxis. And for so many fields, as you gracefully show, they have eliminated our centuries old method for occasioning and evaluating learning.
I’ve moved to Socratic dialogue, oral examinations, and disputations—not because I have given up on writing as essential, but precisely to help students encounter the “cognitive pushback” in their learning. It requires them to represent the level of their preparation themselves (or face up to the reality that they didn’t prepare). They and their performance in front of their peers becomes one of those “friction points” that helps them, at the very least, orient themselves to how competent they are becoming.
I have final paper conferences in my smaller classes, which serve some of these functions. It’s nice to be able to sit and talk about a paper that they’ve poured a lot of effort into rather than just leaving marginal notes. I haven’t tried a Socratic dialogue, but my kids have had them in their high school and take them very seriously.
I teach English in the European context. I am adamant about students learning to write themselves not just outsourcing it to someone or something else.
In a recent conference including other langauge teachers reflecting on our university programs, I was the only person still teaching writing because everyone else felt they could no longer evaluate student writing. I believe we learn what we think as we write, and this is the reason I keep looking for ways to help students write in the time of AI.
This semester with my first year students, the first few sessions were theme-setting and group-building. Students were assigned readings, taught how to ask an AI to abridge the readings if necessary to an individual’s level + a bit, and in the next class, asked to discuss their readings after everyone moved their chairs into a circle.
The first session, about half the students had done the readings. The second, everyone had. When we got ready to write, students formed peer-review groups to discuss their ideas for answering the prompt. They then went off to write on their own.
In the following class, they shared their writing in peer-review groups, went through a series of questions, and gave one another input. Then they submitted their writing. They had three genres to write.
Their captstone type writing was to think of something within our theme that resonated with them, and to write about it for Substack. Everyone had to create an account, they had to write a draft article, share me a link, and then it was up to them if they publish. No list of questions. No more guidelines. The hope is that I’d scaffolded them by both my input and peer-review (trying to introduce the notion of affinity groups) and now they can find their own voice to write something respecting our theme and of importance to them.
Some students wrote in French and had their worked translated. That I explained was not going to work because we wanted to learn to be as capable in English as possible. Others wrote with AI from the inception. To these students I explained that the whole point was to give them power. The power of the written word.
Many students embraced this work, one sharing that it was unusual to be given so few instructions. Others said it was so hard to work without more explicit instructions.
At this point, I wish I had them longer to get into the discussion about interacting with AI. Once they’ve reached this point, they can better reflect on what it would entail to write for themselves and use an AI as an interactive tool.
I appreciate this perspective and I find myself in a similar situation. I have tried to orchestrate what the researchers you cited called “effortful learning” in my job as a professor of religious studies by applying what I learned when I taught junior high science. Inquiry based learning is more effective and longer lasting than lecture.
The problem I’m finding, increasingly, is that students are resistant to this type of learning. I teach required courses. I always have to get the students on board somehow. In the past, it’s been easier. They’ve been more willing to play with ideas, to sit in difficulty. Now, if I don’t give them a straight answer, they just sort of shut down. I’m not sure if it’s because they don’t care or if they just don’t know how to play with ideas. This doesn’t strike me as a problem Bloom can solve, at least the cognitive domain of Bloom.
So, in addition to thinking about how to reframe my pedagogy, I’m also asking myself about whether, if at all, I can help cultivate student motivation. Or, is there a dimension to the professor-student relationship that I need to attend to?
I've definitely experienced shifts in what students resist, too. I've been paying a lot more attention to including community-building activities. At first it felt like time lost, but for me it has paid off pretty consistently. I've seen classes that are reluctant to talk even in small groups at the start of the semester become bubbly and friendly and chatty with each other within a few weeks--and once they have that rapport they seem much more willing to play with ideas and wrestle with ambiguity. It's not a magic bullet, but it helps enough that I can confidently recommend it as something to be deliberate about. Are there certain kinds of inquiry-based activities that seem to trigger the shutdown response more than others? Or does it feel like a more a general shift?
Thanks for this reminder to do the communty-building activities early in the semester. As someone with a melancholic temperament I *hate* them, but I do find they work. And these post-COVID kids do need (and welcome, I think) help socializing.
I need to think about what precipitates the shutdowns. I would say it's certainly (a) any time I ask a question that requires higher-order thinking—something that isn't directly in the notes. It's also (b) general brainstorming. For example, in a unit on religious disaffiliation, students were much less inclined this semester to offer suggestions about why people might leave a religious tradition. Finally, it's (c) controversial topics and (d) personal topics—things that are perceived either as taboo or as so subjective they don't require sustained, systematic discussion. This is tricky with religion. I go back and forth about how to address this.
All in all, I think I need to spend more time intentionally planning discussion—terrifying, since it is so unpredictable, even with planning!
We've lost the ability to project classwork into students' time at home, and mourn that loss -- and are reluctant to accept that and struggle to internalize it. We can keep doing what we've been doing, to the benefit of students still willing to do that work, or give up entirely and rely on in-class assessment during the time we have each week, or radically expand class time. That last one will meet tremendous resistance as it'll mean cloistering students for long stretches, but will only seem like an overreaction to anyone who hasn't come to terms with the magnitude of the problem. (Call me an alarmist but we alarmists are having our day.)
I don't think that's alarmist at all--it's just a description of what has happened. There are lots of things we're going to have to figure out as a result, but at the top of the list for me is figuring out what sorts of work--real, time-consuming, difficult, learning-producing work--we can give them to do outside of class. (Or, rather, meaningful work that we can give them and also have some reliable sense of whether they've done it and what they've learned from it.)
I think what makes this diagnosis feel alarmist is that we can obviously still give them the same meaningful work we've always given them (reading, writing, problem sets, etc), and *they can still do that work* and derive enormous benefits. The vast majority of students (or at least of my students) are still doing the work because their purpose in college is to gain a real education. It's what they want, so it's what they do. And as long as that's true, it feels alarmist to freak out.
But across higher ed, it increasingly feels like that norm is eroding. Many of my grad school friends, who teach at very different places, are downright despondent. The truth is that we've never been able to *make* anyone do the work, but there have been fairly reliable signals we can look for to see if they've done it and to assess what they've learned. The carrot-and-stick of grades was usually enough. It's not alarmist to point out that students no longer have to do the work we assign them to have something to turn in, and aside from gut-level suspicions or devoting serious class time to seeing what they know, it's much harder than ever to tell what they have and haven't done or learned.
Sorry for the long response, but I think you hit the nail on the head. What we need is new ways "to project classwork into students' time at home." The authentic work has always been "Read this. Write that. Work through this tricky problem." Not being able to ask them to do those things without a helpful chatbot popping up on the screen to ask if it can do it for them is heartbreaking. It might be great for "productivity," but it's a nightmare for traditional teaching and learning.
Our current class design assumes that students read and write at home. The humanities need to add on a “lab” component that gives us oversight on student work. It’s not just the writing but the reading that is being eroded (since students can upload pdf files).
Yes, I'm hearing more and more calls for something like this, and I can see some real advantages to having more time to do meaningful things together in class. (I also have some fears. I can't help but think about old-fashioned "study hall", the kind of labor arrangements we'd have to make to implement labs, and sending the signal that learning is something that can only be done with oversight vs in an independent, self-directed sort of way. None of these are insurmountable and the benefits might outweigh the costs. Until people try it we're all just guessing. But I still worry.)
I'm out of the undergrad teaching game for a few years but I've imagined revamped reading quizzes thare are so nickpicky that no AI-generated summary will suffice. Along the lines of, "Which of the following sentences was/was not in the reading?" But I'm reluctant to implement any change that will be hated by students without course evaluation forgiveness. Promised in advance, in writing.
I see the breakdown happening earlier. Building sentences and choosing words to imperfectly match meaning is fundamental to developing critical thinking skills. When my students realized how much they were missing — which they did when I taught “how to write a sentence” to undergrads, they mourned and celebrated at the same time. Why we are robbing them of this experience and exercise is truly beyond me.
Yes! Writing well and thinking critically (not to mention reading deeply) are all foundational and worth fighting for. I think that's why I feel so much grief for the way that AI is undercutting the traditional college essay. I'm in the lucky position that the vast majority of my students still would rather labor over their writing and learn something than use AI as a way to escape the work. But "vast majority" doesn't mean 100%, and right now I worry that what's protecting their learning is their motivation rather than anything special about what I'm asking them to do. Have you found any structures that help enforce the work aside from (in addition to?) giving them something to celebrate?
I'm interested in hearing what sort of alternate assignments educators are developing or pivoting to in light of this issue with AI. I've heard a lot about returning to blue books, but I feel like that will miss the students who don't test well -- which was why education had moved beyond them in the first place. What other options are there, old or new?
(For the record -- I currently teach outdoor school, and so I'm blessedly free of this debate at this point in my career. AI can't help you figure out how to work as a team in the middle of the woods ;) But I'm still paying attention, and interested in what this dilemma illuminates about the entire endeavor of teaching -- and learning.)
Thanks, KLB. I've also heard a lot about experiments with blue books, oral exams, presentations, and other more traditional testing methods. I know some people have had good success with that, but I share your equity concerns. Even before AI hit, I'd shifted during COVID to a combination of specifications ("specs") grading in big classes (regular reading responses, unit essays, etc) and in small classes stuck with traditional papers but added paper conferences (20-30 minutes per student with discussion and oral feedback from me). It's still mostly working--but the essay's opening anecdote is from one of the specs based classes. I'm actually wondering if what you're doing to get them to work as a team in the middle of the woods might have some lessons for us rather than the other way around! Anything you're seeing in places where all the screens are off that works really well?
In smaller seminar courses I've seen moderate success with forging personal connections with students, no- tech classrooms, oral exams, bluebooks, and an emphasis on creative material projects that require hands-on involvement (making an artifact, cooking, etc). But for large lectures & especially for online courses nobody seems to have good answers.
That matches up pretty directly with what I'm seeing/hearing as well.
I think Anthropic, OpenAI et al should pay for a luxe summer camp for faculty invested in these questions.
In reality I think they owe Education so much more. They're disrupting an industry with which they are not in economic competition, as say Uber : Taxis. And for so many fields, as you gracefully show, they have eliminated our centuries old method for occasioning and evaluating learning.
I’ve moved to Socratic dialogue, oral examinations, and disputations—not because I have given up on writing as essential, but precisely to help students encounter the “cognitive pushback” in their learning. It requires them to represent the level of their preparation themselves (or face up to the reality that they didn’t prepare). They and their performance in front of their peers becomes one of those “friction points” that helps them, at the very least, orient themselves to how competent they are becoming.
I have final paper conferences in my smaller classes, which serve some of these functions. It’s nice to be able to sit and talk about a paper that they’ve poured a lot of effort into rather than just leaving marginal notes. I haven’t tried a Socratic dialogue, but my kids have had them in their high school and take them very seriously.
I teach English in the European context. I am adamant about students learning to write themselves not just outsourcing it to someone or something else.
In a recent conference including other langauge teachers reflecting on our university programs, I was the only person still teaching writing because everyone else felt they could no longer evaluate student writing. I believe we learn what we think as we write, and this is the reason I keep looking for ways to help students write in the time of AI.
This semester with my first year students, the first few sessions were theme-setting and group-building. Students were assigned readings, taught how to ask an AI to abridge the readings if necessary to an individual’s level + a bit, and in the next class, asked to discuss their readings after everyone moved their chairs into a circle.
The first session, about half the students had done the readings. The second, everyone had. When we got ready to write, students formed peer-review groups to discuss their ideas for answering the prompt. They then went off to write on their own.
In the following class, they shared their writing in peer-review groups, went through a series of questions, and gave one another input. Then they submitted their writing. They had three genres to write.
Their captstone type writing was to think of something within our theme that resonated with them, and to write about it for Substack. Everyone had to create an account, they had to write a draft article, share me a link, and then it was up to them if they publish. No list of questions. No more guidelines. The hope is that I’d scaffolded them by both my input and peer-review (trying to introduce the notion of affinity groups) and now they can find their own voice to write something respecting our theme and of importance to them.
Some students wrote in French and had their worked translated. That I explained was not going to work because we wanted to learn to be as capable in English as possible. Others wrote with AI from the inception. To these students I explained that the whole point was to give them power. The power of the written word.
Many students embraced this work, one sharing that it was unusual to be given so few instructions. Others said it was so hard to work without more explicit instructions.
At this point, I wish I had them longer to get into the discussion about interacting with AI. Once they’ve reached this point, they can better reflect on what it would entail to write for themselves and use an AI as an interactive tool.
This sounds fantastic! Thanks for sharing—I especially like the idea of something ready to go in Substack, with publication being their call.
Spot on. The product was once valid evidence of the process but now you can skip straight to product as if the product was the only point.
I appreciate this perspective and I find myself in a similar situation. I have tried to orchestrate what the researchers you cited called “effortful learning” in my job as a professor of religious studies by applying what I learned when I taught junior high science. Inquiry based learning is more effective and longer lasting than lecture.
The problem I’m finding, increasingly, is that students are resistant to this type of learning. I teach required courses. I always have to get the students on board somehow. In the past, it’s been easier. They’ve been more willing to play with ideas, to sit in difficulty. Now, if I don’t give them a straight answer, they just sort of shut down. I’m not sure if it’s because they don’t care or if they just don’t know how to play with ideas. This doesn’t strike me as a problem Bloom can solve, at least the cognitive domain of Bloom.
So, in addition to thinking about how to reframe my pedagogy, I’m also asking myself about whether, if at all, I can help cultivate student motivation. Or, is there a dimension to the professor-student relationship that I need to attend to?
I've definitely experienced shifts in what students resist, too. I've been paying a lot more attention to including community-building activities. At first it felt like time lost, but for me it has paid off pretty consistently. I've seen classes that are reluctant to talk even in small groups at the start of the semester become bubbly and friendly and chatty with each other within a few weeks--and once they have that rapport they seem much more willing to play with ideas and wrestle with ambiguity. It's not a magic bullet, but it helps enough that I can confidently recommend it as something to be deliberate about. Are there certain kinds of inquiry-based activities that seem to trigger the shutdown response more than others? Or does it feel like a more a general shift?
Thanks for this reminder to do the communty-building activities early in the semester. As someone with a melancholic temperament I *hate* them, but I do find they work. And these post-COVID kids do need (and welcome, I think) help socializing.
I need to think about what precipitates the shutdowns. I would say it's certainly (a) any time I ask a question that requires higher-order thinking—something that isn't directly in the notes. It's also (b) general brainstorming. For example, in a unit on religious disaffiliation, students were much less inclined this semester to offer suggestions about why people might leave a religious tradition. Finally, it's (c) controversial topics and (d) personal topics—things that are perceived either as taboo or as so subjective they don't require sustained, systematic discussion. This is tricky with religion. I go back and forth about how to address this.
All in all, I think I need to spend more time intentionally planning discussion—terrifying, since it is so unpredictable, even with planning!