I see the use for this model in response to writing assignments for 9-12 US English. Once I’ve defined the objectives (NOT a grading rubric), the app’s response to what the student writes can come as Socratic questions (NOT a grade). I’d like to use the method for revision, IF I could develop an app that would let me read the essays and the chats. Humanly, finding time to do/remember all that reading is a design challenge. But it might work really well in a think/pair/share mode.
Perhaps unsurprisingly, I don't love it. In many ways, educators in the K12 space have been curating their own curriculum in similar fashion using sites such as Teacher Pay Teachers. That site and other like it are a wasteland of terrible lessons and task ideas that nonetheless get can great traction for all the wrong reasons (easy to implement, students may enjoy them, but no learning takes place). It's not clear to me that AI-produced materials are substantially better; my trial usage suggests they are often worse.
I think it's important to be careful not to "optimize" around what the most thoughtful teachers will do. The real test is how it cashes out with the practical day-to-day realities. And I've seen versions of this movie before.
I guess my question to you would be: how do you see this connected to the question of how much autonomy/trust individual teachers should be given?
For example, in my classroom, I can teach almost whatever I want, almost however I want to. In a situation like this, I am almost completely free to develop my own curriculum as I would like. In a situation like this, I have a vision of what I want my students to do and learn, and the learning outcomes for my courses.
Perhaps I am reading you incorrectly here (and please correct me if I am), but my sense is that you are operating with a sense of distrust in the individual teacher in the classroom and questioning (at minimum) their ability to choose and implement curriculum that is fruitful for their students (e.g., your Teachers Pay Teachers vignette).
I think there is a fine balance here where we have a set of curriculum and materials that are solid foundations, but the joy and art of teaching is often found in finding your own ways to put a spin on that material in order to help it connect with your students and to target specific misconceptions. I would argue that almost all of the best teachers do this. At least this bears out in my experience.
Perhaps your point is more about specific age ranges where this approach may be more or less valid (e.g., more applicable at the collegiate level, less applicable in 3rd grade), but I am curious to hear more about the core concern here.
I'm certainly not arguing that we discard all our curriculum and replace it with AI-generated material. But given my own experience, I'm bullish on the ability of AI to help me create new ways of visualizing and engaging with that curriculum that will help students to understand it more fully.
Appreciate the thoughtful response to my comment. A few more thoughts from moi:
1. It's not so much about distrust or trying to limit teacher autonomy that motivated me to respond here. Sites like Teachers Pay Teachers exist, and I am not advocating for banning them, nor AI (though I'm becoming abolitionist-curious). And I fully agree that we want teachers to find joy in how they approach their instruction.
The issue I have here is centering the use of AI in the mental model of teaching you're promoting. Equating "forward deployed" -- the phrase used to be "future focused" -- with the use of technology is precisely what's landed us in this mess, it further amplifies the ideology of technology, as Ursula Franklin would put it.
2. Another big loss from the approach you advocate here is the hit to collective intelligence. This is more of a thing in K12, but when there's not a shared curricular foundation within a school (or school district), it's difficult for teachers to collaborate and support each other, both within subjects and across grade levels. We know from cognitive science that our ability to understand new ideas depends on ideas we already know, so sequencing the building of knowlege -- not just in a single course, but over the span of schooling -- is vital. We are absolutely terrible at this in the US.
In Japan, by contrast, you can actually see the same lesson being taught on the same topic on the same day throughout the country. BEcause of this, there is intensive shared professional development -- "Japanese lesson study" -- that teachers undertake together to improve their pedagogy. In a perhaps related story, on the int'l PISA exam Japan ranks 2nd, 3rd, and 5th on science, math, and reading, respectively.
So I don't think we need "Forward Deployed Educators," sorry. I think we need more socially connected humans, working and learning together.
I enjoyed your framework a lot. I've been using Claude to help me create artifacts for my younger students as part of our homeschool curriculum. My son and I are starting Alice in Wonderland, so I asked Claude to help me build an interdisciplinary deep reading course, centered around his love of math, puzzles, and photography. We did maybe 3 rounds of back and forth before I felt really good about what we created. I'm excited to try it with him now.
Having tools like this feels like a superpower for creating a really deep education, even at a young age.
I see the use for this model in response to writing assignments for 9-12 US English. Once I’ve defined the objectives (NOT a grading rubric), the app’s response to what the student writes can come as Socratic questions (NOT a grade). I’d like to use the method for revision, IF I could develop an app that would let me read the essays and the chats. Humanly, finding time to do/remember all that reading is a design challenge. But it might work really well in a think/pair/share mode.
Perhaps unsurprisingly, I don't love it. In many ways, educators in the K12 space have been curating their own curriculum in similar fashion using sites such as Teacher Pay Teachers. That site and other like it are a wasteland of terrible lessons and task ideas that nonetheless get can great traction for all the wrong reasons (easy to implement, students may enjoy them, but no learning takes place). It's not clear to me that AI-produced materials are substantially better; my trial usage suggests they are often worse.
I think it's important to be careful not to "optimize" around what the most thoughtful teachers will do. The real test is how it cashes out with the practical day-to-day realities. And I've seen versions of this movie before.
Not surprised :)
I guess my question to you would be: how do you see this connected to the question of how much autonomy/trust individual teachers should be given?
For example, in my classroom, I can teach almost whatever I want, almost however I want to. In a situation like this, I am almost completely free to develop my own curriculum as I would like. In a situation like this, I have a vision of what I want my students to do and learn, and the learning outcomes for my courses.
Perhaps I am reading you incorrectly here (and please correct me if I am), but my sense is that you are operating with a sense of distrust in the individual teacher in the classroom and questioning (at minimum) their ability to choose and implement curriculum that is fruitful for their students (e.g., your Teachers Pay Teachers vignette).
I think there is a fine balance here where we have a set of curriculum and materials that are solid foundations, but the joy and art of teaching is often found in finding your own ways to put a spin on that material in order to help it connect with your students and to target specific misconceptions. I would argue that almost all of the best teachers do this. At least this bears out in my experience.
Perhaps your point is more about specific age ranges where this approach may be more or less valid (e.g., more applicable at the collegiate level, less applicable in 3rd grade), but I am curious to hear more about the core concern here.
I'm certainly not arguing that we discard all our curriculum and replace it with AI-generated material. But given my own experience, I'm bullish on the ability of AI to help me create new ways of visualizing and engaging with that curriculum that will help students to understand it more fully.
Appreciate the thoughtful response to my comment. A few more thoughts from moi:
1. It's not so much about distrust or trying to limit teacher autonomy that motivated me to respond here. Sites like Teachers Pay Teachers exist, and I am not advocating for banning them, nor AI (though I'm becoming abolitionist-curious). And I fully agree that we want teachers to find joy in how they approach their instruction.
The issue I have here is centering the use of AI in the mental model of teaching you're promoting. Equating "forward deployed" -- the phrase used to be "future focused" -- with the use of technology is precisely what's landed us in this mess, it further amplifies the ideology of technology, as Ursula Franklin would put it.
2. Another big loss from the approach you advocate here is the hit to collective intelligence. This is more of a thing in K12, but when there's not a shared curricular foundation within a school (or school district), it's difficult for teachers to collaborate and support each other, both within subjects and across grade levels. We know from cognitive science that our ability to understand new ideas depends on ideas we already know, so sequencing the building of knowlege -- not just in a single course, but over the span of schooling -- is vital. We are absolutely terrible at this in the US.
In Japan, by contrast, you can actually see the same lesson being taught on the same topic on the same day throughout the country. BEcause of this, there is intensive shared professional development -- "Japanese lesson study" -- that teachers undertake together to improve their pedagogy. In a perhaps related story, on the int'l PISA exam Japan ranks 2nd, 3rd, and 5th on science, math, and reading, respectively.
So I don't think we need "Forward Deployed Educators," sorry. I think we need more socially connected humans, working and learning together.
What a fascinating concept! This model could be used in so many sectors to improve innovation and adaptation of capabilities.
I enjoyed your framework a lot. I've been using Claude to help me create artifacts for my younger students as part of our homeschool curriculum. My son and I are starting Alice in Wonderland, so I asked Claude to help me build an interdisciplinary deep reading course, centered around his love of math, puzzles, and photography. We did maybe 3 rounds of back and forth before I felt really good about what we created. I'm excited to try it with him now.
Having tools like this feels like a superpower for creating a really deep education, even at a young age.
I really appreciated this article and your diagram. Thank you.
Thank you!