College diplomas and even GPAs are still the gatekeepers for new hires, though in some industries we are seeing a shift. Did universities talk employers into valuing those? Can universities talk employers out of valuing those? Maybe it’s the only game in town. Even if academia managed to change the educational process, the shingle would still be a gatekeeper. But I hear you. I dare to hope that, when I tell my students that it will be more valuable to learn in my course instead of getting the next higher letter grade, some of them will actually believe me and act on my assertion.
That’s a great point. I wonder — and I legitimately don’t know — if the gatekeeping function of grades is changing even without our work. We have the (arguable) death of the “college for all” movement, grade creep, and other things that have hurt the validity of grades on a broad level even as just markers of achievement. I could see that opening up a path for something else, though that could be me being overly optimistic.
None of these will work. The transactional nature of education is pulled from the educational system, not pushed. Reduce demand for grades and degrees and students will stop seeing school, particularly higher ed, as transactional.
I definitely think the reduced demand for grades and degrees — as well as a gradual end to the “college for all movement” that peaked during the Obama admin — will play a huge part.
I’m not sure if I get the point that these sorts of initiatives will fail, because this transactional model comes from higher ed. Is there much evidence to support that?
Most of the genealogies I’ve seen focus primarily on the post-WWII government initiatives that incentivized college and seemed to promote a transactional model. (This is when the A,B,C mode became entrenched in education, as well.) It doesn’t seem like college education was quite as transactional (at least not broadly) until around that time.
I’m definitely open to other genealogies, especially if they show the transactional model as stemming from within higher ed.
There does seem to be quite a bit of evidence that these initiatives — on a class-basis as well as a college-basis — work quite well, and pretty immediately.
Part of the process of leaving behind transactional education is knowing what you’re leaving it FOR—what your destination is in mind. Why are you educating the kid in front of you? Is this a writing class? Then not learning to write is a fail, obviously. But how do you SHOW writing? Mainly by doing. Writing is a slow and methodical process of gathering sources and evaluating them and ranking them and presenting them with connective phrases and ideas to make those sources convincing. You can’t convince someone by throwing a paper at them, you have to persuade them. So what is a writing class? A persuading class. Prove to me that you can do this work. In the age of AI, how do you persuade someone? In person, speaking to them like real people. It’s how we do dissertation defenses, it should be how we do all papers. If you can get ONE good, well-argued and successfully defended paper out of a student after a whole semester, they should get an A. Maybe a B if they can’t really argue their points well, because that’s a public speaking issue.
That’s how we educate on process—flip the script on the student. Instead of pouring info into vapid minds, start the very first class with the end assignment. “I want you to prove to me that you can write well and defend your ideas. Who’s first?” Give the students free rein to use any chat bots they want, and let them present their papers to the class. Go over their sources in class. Show where those sources are fake or don’t claim what the chatbox says. Interrogate the usefulness of the source, make the student prove their worth and say that their ability needs to be demonstrated. This is not a thought exercise, this is a learned skill. And if you want to shortcut that skill, it will show. And if you get a good student who happens to use a chat box to get what they need and then edits and corrects and expands where you want them to, congrats—you’ve learned to write by revising. It’s also an essential skill.
From then on, the student will instinctively know that no section of the paper will be left unscrutinized, because these will be read out loud and discussed in class. It’s the way the British system does it at Oxford, it should be the standard in the US as well.
I can’t speak for too much of Oxford nowadays, but I was there in the 00s and dons required weekly papers on the subjects we were interested in. The British university system is not a liberal arts education, lots of different classes about a wide range of topics. It’s targeted, because having your A-levels assumes that you already did the broad education. Closer to a graduate degree, but we also overload broad learning and language skills even in grad schools.
Frankly, the American education system re-teaches a lot of information that it should have sufficed to learn once. I personally think it’s the lecture-based method that doesn’t do a good job of sticking: who is going to retain even a fraction of information when the learning is 90% passive with only 10% application? Doing anything more seems like work, and students don’t come to class to work, they come to class to learn. I flipped the script in one class and had no set lectures, just paper writing all semester. One student asked if he could just stay home then, since there wasn’t any lecture. I said no—how do you learn with nobody to guide you in the room? Turns out, writing for two classes a week is HARD, and it requires a lot of the teacher. It wasn’t a huge success, but boy, it was a different experience. More like a workshop than a “class” class.
Well written piece Jason. The transactional nature also adds to the power base. Many teachers enjoy having this power to be able to tell students what to do. Giving it up, letting students evaluate their own work is something that removes this type of power. Yet kids can always tell you whether they did well or not. One of the other issue is testing itself. It’s not teaching or learning, it’s a time when neither of these things happen. Its purpose is not for the student it’s usually for the system, the transactional system as you put it. Ciao Cathy
These are great points. I think that alternative assessment and similar techniques really does redistribute power, at least partially, and it makes a big difference.
This was a huge problem before AI, AI just helps us see it clearer I think. If we could at least get students to think in terms of learning and competency in their chosen field of study, that would be a huge step forward. Sure, the "core classes" are supposed to help create well-rounded humans, and I support that idea, but I could forgive a student for just getting by to get through in a course they just had to take that wasn't really in their area of focus.
I think the big thing is that evidence suggests that learning-centered classrooms are better for skill acquisition. And it seems like that’s becoming the differentiator, especially as the trade-in value of a college degree has decreased.
I wonder if generative AI will cause (along with other factors) a break between transactional credentialling and learning, such that we won't be chipping away but abandoning credentialling machines to create spaces and experiences that prioritize learning.
Even in middle school, students ask "Is this graded?" Meaning do I have to do my best work here. I give a little of things for learning only like you suggest--and I agree, we need to do more of this to shift the culture.
I liked your piece, Jason. I was surprised to find that many of your suggestions are already things I do in my classes, but I hadn't realized that they might be synergizing with AI. For example, I use a rating system I call ESPN (Excellent, Satisfactory, Progressing, Not submitted) instead of points. It’s essentially glorified pass/no-pass: S = A-level mastery, while P is anything below that. E is reserved for outstanding work and is rare, like a carrot at the end of the stick.
The key is that students can resubmit as often as they want; most hit mastery by draft two, and never past four. So “P” means "the ball’s in your court, your move", not a penalty. Of course, I use AI to help triage feedback—that’s how I can afford “infinite” resubmissions!
I also hold letter grades until the end. Some students hate this; others breathe more easily. The odd outcome is that when mastery is the bar and iteration is normal, A grades aren’t scarce. Most get there, and having earned it, not on a curve.
Not sure that’s enough, though. I reinforce it by asking for a two-sentence reflection with each resubmission (what changed and why). We finish the course with a short oral exam, announced on day one, as a motivator, not a trap: students know they’ll have to explain their work live, so it’s more economical to plan early than cram.
I’m curious about your thoughts: do delayed letter grades help learning or just move the anxiety? I’m still on the fence, though it still feels better than the usual alternative. Also, how do you balance transparency without turning everything into points?
This is why I’m so excited about relational AI. It harnesses the generativity of Gen AI, to augment and a company, rather than replace the thinking of students, and really anyone else who engages with it relationally. If we get away from the idea that we’re talking to a person, and we approach it as talking to information itself, there’s so much more that we can do with it, that the transactional model can’t even fathom. It’s not gen AI that’s the problem, it’s the approach. And the opportunities are pretty amazing.
Could not agree more. I've been saying we need to move from a transactional to a learning mindset for awhile now. The big question is how.
https://aigoestocollege.substack.com/p/beyond-grades-transitioning-from
Absolutely! And I’ll check out your post. Sorry if I missed it!
College diplomas and even GPAs are still the gatekeepers for new hires, though in some industries we are seeing a shift. Did universities talk employers into valuing those? Can universities talk employers out of valuing those? Maybe it’s the only game in town. Even if academia managed to change the educational process, the shingle would still be a gatekeeper. But I hear you. I dare to hope that, when I tell my students that it will be more valuable to learn in my course instead of getting the next higher letter grade, some of them will actually believe me and act on my assertion.
That’s a great point. I wonder — and I legitimately don’t know — if the gatekeeping function of grades is changing even without our work. We have the (arguable) death of the “college for all” movement, grade creep, and other things that have hurt the validity of grades on a broad level even as just markers of achievement. I could see that opening up a path for something else, though that could be me being overly optimistic.
None of these will work. The transactional nature of education is pulled from the educational system, not pushed. Reduce demand for grades and degrees and students will stop seeing school, particularly higher ed, as transactional.
Thanks, Brian!
I definitely think the reduced demand for grades and degrees — as well as a gradual end to the “college for all movement” that peaked during the Obama admin — will play a huge part.
I’m not sure if I get the point that these sorts of initiatives will fail, because this transactional model comes from higher ed. Is there much evidence to support that?
Most of the genealogies I’ve seen focus primarily on the post-WWII government initiatives that incentivized college and seemed to promote a transactional model. (This is when the A,B,C mode became entrenched in education, as well.) It doesn’t seem like college education was quite as transactional (at least not broadly) until around that time.
I’m definitely open to other genealogies, especially if they show the transactional model as stemming from within higher ed.
There does seem to be quite a bit of evidence that these initiatives — on a class-basis as well as a college-basis — work quite well, and pretty immediately.
Part of the process of leaving behind transactional education is knowing what you’re leaving it FOR—what your destination is in mind. Why are you educating the kid in front of you? Is this a writing class? Then not learning to write is a fail, obviously. But how do you SHOW writing? Mainly by doing. Writing is a slow and methodical process of gathering sources and evaluating them and ranking them and presenting them with connective phrases and ideas to make those sources convincing. You can’t convince someone by throwing a paper at them, you have to persuade them. So what is a writing class? A persuading class. Prove to me that you can do this work. In the age of AI, how do you persuade someone? In person, speaking to them like real people. It’s how we do dissertation defenses, it should be how we do all papers. If you can get ONE good, well-argued and successfully defended paper out of a student after a whole semester, they should get an A. Maybe a B if they can’t really argue their points well, because that’s a public speaking issue.
That’s how we educate on process—flip the script on the student. Instead of pouring info into vapid minds, start the very first class with the end assignment. “I want you to prove to me that you can write well and defend your ideas. Who’s first?” Give the students free rein to use any chat bots they want, and let them present their papers to the class. Go over their sources in class. Show where those sources are fake or don’t claim what the chatbox says. Interrogate the usefulness of the source, make the student prove their worth and say that their ability needs to be demonstrated. This is not a thought exercise, this is a learned skill. And if you want to shortcut that skill, it will show. And if you get a good student who happens to use a chat box to get what they need and then edits and corrects and expands where you want them to, congrats—you’ve learned to write by revising. It’s also an essential skill.
From then on, the student will instinctively know that no section of the paper will be left unscrutinized, because these will be read out loud and discussed in class. It’s the way the British system does it at Oxford, it should be the standard in the US as well.
This is a great point, and I really like this system and approach.
Is this a widespread approach at Oxford? Does it cut across departments or is this just in the writing disciplines?
I can’t speak for too much of Oxford nowadays, but I was there in the 00s and dons required weekly papers on the subjects we were interested in. The British university system is not a liberal arts education, lots of different classes about a wide range of topics. It’s targeted, because having your A-levels assumes that you already did the broad education. Closer to a graduate degree, but we also overload broad learning and language skills even in grad schools.
Frankly, the American education system re-teaches a lot of information that it should have sufficed to learn once. I personally think it’s the lecture-based method that doesn’t do a good job of sticking: who is going to retain even a fraction of information when the learning is 90% passive with only 10% application? Doing anything more seems like work, and students don’t come to class to work, they come to class to learn. I flipped the script in one class and had no set lectures, just paper writing all semester. One student asked if he could just stay home then, since there wasn’t any lecture. I said no—how do you learn with nobody to guide you in the room? Turns out, writing for two classes a week is HARD, and it requires a lot of the teacher. It wasn’t a huge success, but boy, it was a different experience. More like a workshop than a “class” class.
Well written piece Jason. The transactional nature also adds to the power base. Many teachers enjoy having this power to be able to tell students what to do. Giving it up, letting students evaluate their own work is something that removes this type of power. Yet kids can always tell you whether they did well or not. One of the other issue is testing itself. It’s not teaching or learning, it’s a time when neither of these things happen. Its purpose is not for the student it’s usually for the system, the transactional system as you put it. Ciao Cathy
These are great points. I think that alternative assessment and similar techniques really does redistribute power, at least partially, and it makes a big difference.
This was a huge problem before AI, AI just helps us see it clearer I think. If we could at least get students to think in terms of learning and competency in their chosen field of study, that would be a huge step forward. Sure, the "core classes" are supposed to help create well-rounded humans, and I support that idea, but I could forgive a student for just getting by to get through in a course they just had to take that wasn't really in their area of focus.
Right!
I think the big thing is that evidence suggests that learning-centered classrooms are better for skill acquisition. And it seems like that’s becoming the differentiator, especially as the trade-in value of a college degree has decreased.
I wonder if generative AI will cause (along with other factors) a break between transactional credentialling and learning, such that we won't be chipping away but abandoning credentialling machines to create spaces and experiences that prioritize learning.
I think that is certainly possible!
Even in middle school, students ask "Is this graded?" Meaning do I have to do my best work here. I give a little of things for learning only like you suggest--and I agree, we need to do more of this to shift the culture.
Right! I think that question shows a lot when we encounter it. And it often happens so early.
Chipping away is the work we all need to be doing.
Right! And I’m hoping the chipping away will show the cracks in the foundation and we can just the whole thing (transactional education) down.
I liked your piece, Jason. I was surprised to find that many of your suggestions are already things I do in my classes, but I hadn't realized that they might be synergizing with AI. For example, I use a rating system I call ESPN (Excellent, Satisfactory, Progressing, Not submitted) instead of points. It’s essentially glorified pass/no-pass: S = A-level mastery, while P is anything below that. E is reserved for outstanding work and is rare, like a carrot at the end of the stick.
The key is that students can resubmit as often as they want; most hit mastery by draft two, and never past four. So “P” means "the ball’s in your court, your move", not a penalty. Of course, I use AI to help triage feedback—that’s how I can afford “infinite” resubmissions!
I also hold letter grades until the end. Some students hate this; others breathe more easily. The odd outcome is that when mastery is the bar and iteration is normal, A grades aren’t scarce. Most get there, and having earned it, not on a curve.
Not sure that’s enough, though. I reinforce it by asking for a two-sentence reflection with each resubmission (what changed and why). We finish the course with a short oral exam, announced on day one, as a motivator, not a trap: students know they’ll have to explain their work live, so it’s more economical to plan early than cram.
I’m curious about your thoughts: do delayed letter grades help learning or just move the anxiety? I’m still on the fence, though it still feels better than the usual alternative. Also, how do you balance transparency without turning everything into points?
This is why I’m so excited about relational AI. It harnesses the generativity of Gen AI, to augment and a company, rather than replace the thinking of students, and really anyone else who engages with it relationally. If we get away from the idea that we’re talking to a person, and we approach it as talking to information itself, there’s so much more that we can do with it, that the transactional model can’t even fathom. It’s not gen AI that’s the problem, it’s the approach. And the opportunities are pretty amazing.