I was reading a foundational document from our national educational research organization called "How students learn" and in it it makes the claim that "learning is a change in long term memory". This supports the notion that having a capacity to recall stuff is foundational to any learning. However, my memory doesn't seem to do this well, yet I've done a lot of learning over the years. Enjoyed this read very much, Karl. I'll do my best to remember some of it.
“AI's capacity for continuous, granular data collection and real-time adaptation could *possibly* allow us to map the precise conditions under which, say, spacing intervals should expand or contract…” - love this. It pushes for me how AI enabled personalisation of learning has been imagined.
"If students rely on AI to generate ideas, structure arguments, or retrieve facts, they run the risk of cosplaying domain knowledge while bypassing the effortful processes that lead to genuine learning." Love it.
I'm conflicted when it comes to AI. I like to use it is as a thinking partner-- much in the way Vannevar Bush describes using the theoretical Memex machine, from his essay "As We May Think."
When it comes to schools, I think that aligning it to the classical trivium could make sense-- as a tool to be accessed later in the K12 experience-- high school.
In the small org I currently lead, we are doing away with 1:1 chromebooks next year and ridding ourselves of those awful online learning programs, BUT we are piloting an AI decoding coach that can highlight, pause and coach students to through decoding errors.
That sounds kind of contradictory to some--maybe it is-- but I don't really see it that way.
Needless to say, I'll be watching the data very closely and will most likely pull students to listen to them read.
After reading your article, I found myself reflecting on the deeper role of knowledge and the potential (and pitfalls) of AI in education. Thank you for such a thought-provoking and well-argued piece.
My Marginalia Notes: Plan B: Bloom (1913-1999) - Mastery learning. Taxonomy of learning not a hierarchy …
The place of AI in learning
A new useful term “taxonomic fallacy”: the mistake of treating the first step in learning (e.g., labeling or remembering) as though it were the final step (understanding or application). True learning involves both, the accumulation of factual knowledge and the ability to revisit, expand, and connect it.
As the article rightly emphasizes:
“In reality, factual knowledge is the substrate from which all higher-order thinking grows. I’ll just say it again: you can’t connect the dots if you don’t have any dots.”
This is an essential point. Remembering is not inferior to analysis or creativity, it’s integral. Learning is not a climb up a pyramid, but a web of connections.
So where does AI fit in?
AI can support personalised learning, by adapting pace, content, and feedback to individual learners. This is where Bloom’s vision of mastery learning, freed from the “tyranny of time”, becomes truly feasible.
But, the article warns, us to be cautious:
“Students may become more sophisticated at prompting and curating AI outputs without developing the underlying disciplinary thinking the AI is ostensibly supporting.”
In other words, without a solid foundation of knowledge and understanding, learners risk outsourcing their thinking to tools, mistaking fluency with AI for actual expertise. AI should amplify independent thought, not replace it.
I really appreciate the Willingham reference, as it supports what I’ve observed in the intro art history classroom and in my own education as first a molecular biologist and then an art historian. You have to have something to think *with* in order to think critically about a subject. It really helps with the argument that AI can do the *lower order* things like summoning facts so we should focus on the *higher order* things like creativity. Really helpful and I will be using it in discussions with students about AI.
And why can’t we get learning styles and VARK to go away??
Thank you for this Carl, I have been struggling to reconcile my volatile relationship with AI and this helped greatly.
It’s a teaching assistant, research assistant, structural and copy editor… and a bunch of other help, but its not the learner, not the creator. Obvious enough, but I suspect many are inadvertently assuming it will be a person.
It might, eventually, look like one but it is not real. It hasn’t got a navel.
I was reading a foundational document from our national educational research organization called "How students learn" and in it it makes the claim that "learning is a change in long term memory". This supports the notion that having a capacity to recall stuff is foundational to any learning. However, my memory doesn't seem to do this well, yet I've done a lot of learning over the years. Enjoyed this read very much, Karl. I'll do my best to remember some of it.
Maybe like getting your name correct. Sorry Carl.
“AI's capacity for continuous, granular data collection and real-time adaptation could *possibly* allow us to map the precise conditions under which, say, spacing intervals should expand or contract…” - love this. It pushes for me how AI enabled personalisation of learning has been imagined.
"If students rely on AI to generate ideas, structure arguments, or retrieve facts, they run the risk of cosplaying domain knowledge while bypassing the effortful processes that lead to genuine learning." Love it.
I'm conflicted when it comes to AI. I like to use it is as a thinking partner-- much in the way Vannevar Bush describes using the theoretical Memex machine, from his essay "As We May Think."
When it comes to schools, I think that aligning it to the classical trivium could make sense-- as a tool to be accessed later in the K12 experience-- high school.
In the small org I currently lead, we are doing away with 1:1 chromebooks next year and ridding ourselves of those awful online learning programs, BUT we are piloting an AI decoding coach that can highlight, pause and coach students to through decoding errors.
That sounds kind of contradictory to some--maybe it is-- but I don't really see it that way.
Needless to say, I'll be watching the data very closely and will most likely pull students to listen to them read.
In my ai learning platforms , I am pushing learners to higher levels of learning based upon Blooms taxonomy. Love the Bloom AI image
Well said.
After reading your article, I found myself reflecting on the deeper role of knowledge and the potential (and pitfalls) of AI in education. Thank you for such a thought-provoking and well-argued piece.
My Marginalia Notes: Plan B: Bloom (1913-1999) - Mastery learning. Taxonomy of learning not a hierarchy …
The place of AI in learning
A new useful term “taxonomic fallacy”: the mistake of treating the first step in learning (e.g., labeling or remembering) as though it were the final step (understanding or application). True learning involves both, the accumulation of factual knowledge and the ability to revisit, expand, and connect it.
As the article rightly emphasizes:
“In reality, factual knowledge is the substrate from which all higher-order thinking grows. I’ll just say it again: you can’t connect the dots if you don’t have any dots.”
This is an essential point. Remembering is not inferior to analysis or creativity, it’s integral. Learning is not a climb up a pyramid, but a web of connections.
So where does AI fit in?
AI can support personalised learning, by adapting pace, content, and feedback to individual learners. This is where Bloom’s vision of mastery learning, freed from the “tyranny of time”, becomes truly feasible.
But, the article warns, us to be cautious:
“Students may become more sophisticated at prompting and curating AI outputs without developing the underlying disciplinary thinking the AI is ostensibly supporting.”
In other words, without a solid foundation of knowledge and understanding, learners risk outsourcing their thinking to tools, mistaking fluency with AI for actual expertise. AI should amplify independent thought, not replace it.
I really appreciate the Willingham reference, as it supports what I’ve observed in the intro art history classroom and in my own education as first a molecular biologist and then an art historian. You have to have something to think *with* in order to think critically about a subject. It really helps with the argument that AI can do the *lower order* things like summoning facts so we should focus on the *higher order* things like creativity. Really helpful and I will be using it in discussions with students about AI.
And why can’t we get learning styles and VARK to go away??
Thank you for this Carl, I have been struggling to reconcile my volatile relationship with AI and this helped greatly.
It’s a teaching assistant, research assistant, structural and copy editor… and a bunch of other help, but its not the learner, not the creator. Obvious enough, but I suspect many are inadvertently assuming it will be a person.
It might, eventually, look like one but it is not real. It hasn’t got a navel.