"What we should be more concerned about is how AI cheats students." Fully agree with your emphasis on what students are losing in the break-neck adoption of AI tools. My husband and I recently attended a talk at the Perimeter Institute on AI and education, and it was the most hopeful perspective that I have come across: https://schooloftheunconformed.substack.com/p/learning-fast-and-slow-why-ai-will
Well done! I hope no AI was hurt in the production of this post.
Learning comes from struggle. It's frustrating, time-consuming, but when the light goes on, you remember it because you've had to work to flip that switch of discovery.
When "learning" is too easy, it ceases to be memorable. It becomes ephemeral and forgettable.
I'm not even sure students are swimming in the shallows here--more like they're dipping toes in the ocean and then moving on, never immersing themselves, never feeling the thrill of waves and currents of knowledge, never getting tossed around but emerging smarter and wiser from the experience.
When I graduated college in 2019, I never would have guessed it was the final year of a pure college experience. Once the last Covid restrictions subsided in 2023, generative AI was already off to the races. It’s all very sad.
Beautifully explained with a great comparison to the Industrial Revolution. It’s wild how the IR also revolutionized not just work but society so I wonder what the AI revolution will do to society.
Some of my readers recommended your post as a follow-up. We have many of the same points, but I make a distinction between two essentially different roles of education: mastering skills (e.g. writing, problem solving, etc), where I agree AI can be detrimental for the reasons you have so eloquently stated, and understanding concepts (e.g. what is the quantum mechanical model of the atom?), where AI is a fantastic tool. That balance between skill acquisition (where AI bad) and concept mastery (where AI good) makes the whole AI-in-schools thing very tricky!
I would definitely agree with footnote 2, which may be pushed further in the context of tertiary study when it becomes simply about needing that 'piece of paper' in order to get the job.
At its most naked, this can be seen, for example, in Australia where degrees are more or less sold to foreign students who may or may not even have proficiency in English. What does the degree mean in such a situation? Some sort of entry token to particular types of job, perhaps.
'Education' is one of the nation's top exports, and this practice began long before A.I. entered the equation. In such a mockery of learning, the use of A.I. to write papers is simply the next logical step.
I feel strongly that the best antidote to all of this is going to require educators to double down on building communities of trust with their students. The temptation of product without process is so strong for our busy and overwhelmed students who are continuously hearing that they need to keep up with the cutting edge or be left behind. The best thing we can do is help to articulate the false promises and ask them to trust us in helping them navigate through the storm. At some point I do think we'll need to ask them to tie themselves to the mast of the ship to endure the journey through the cacophony of the AI sirens' calls.
But the fact is not the product of labor but labor itself as an action
I fully agree with your passage about the "paper" being only a proxy of understanding and that the real gain is in the journey not really in the destination.
That has been my most contentious point regarding all the AI hype.
Regarding education, I think the question goes deeper than AI discouraging learning. As I see it, the biggest crisis in education right now is that it cannot state a coherent narrative of what Education is for
This is a question that all cultures must answer, preferably consciously.
Today, the answer is often that education is an economic leverage, meaning being a good consumer. Or as you pointed, a customer
This is the most important Substack post I’ve read in weeks. Thank you for laying out so clearly the skill paths that people go down when they adopt a new technology. Your powers of explanation are on full display here. Bravo!
PS One of my concerns is that people may acknowledge that students are learning less but they will claim that learning is no longer as relevant as it once was because they will point to how students are (seemingly) solving so many new problems and advancing humanity’s knowledge with the help of AI. I witnessed the same thing with digital humanities projects: lots of cool ways to display information using graphics and webpages, but it was never clear if anyone was learning much that was new or important.
I find stark parallels between education (teaching) and the trucking industry (the largest first time employer in the USA). Technology doesn't at first replace workers, rather it competes with them. It creates lower wages, higher turnover, precarity and de-regulation. As workers become transcient laborers, it then becomes much easier to replace them, so disempowered they've become. We see this in education. No lack of teachers, just lack of teachers willing to do the job for peanuts, no respect and the stress/work to having a life ratio. The real issue isn't replacing human labor - it is a social fabric that doesn't distribute the gains to the working class. AI should allow us to pay teachers more, free up for more time on real learning/interaction but rather the opposite occurs because we focus on economic value, not social value.
I believe ne of the biggest failures in our schools during this technological age is that we handed out tablets where students learned only to swipe and tap. We neglected to teach computer skills like typing & using basic software. Imagine AI is available to you and you don't know how to properly type. I wrote more about that here:
Your excellent article moved me to restack it several times with the perspective that the big tech oligarchs are drug pushers. I hope that "junkie" perspective adds value to this whole, critical debate.
Ps I'm new here on substack so it will be much appreciated if you can also support me e.g. by restacking my restacks to your audience...
Interesting article. I admit it raises valid concerns about AI potentially diminishing learning through automation. However, I can't help noticing that it mistakenly assumes essay writing is essential for developing critical thinking, overlooking numerous other effective learning pathways. To illustrate this point, let me rephrase the tenth paragraph as it would have sounded a few decades ago, when similar fears arose when typewriters replaced handwriting:
“Typewriters enable students to produce text without doing the careful work of handwriting. Rather than meticulously crafting each letter by hand, students press keys effortlessly. Rather than carefully constructing elegant calligraphy strokes that demonstrate discipline and attention to detail, they now produce text mechanically. And rather than expressing (and refining) their thoughts through the deliberate, artistic act of penmanship, they quickly type out a first draft or even a final one.”
Just as handwriting is no longer central yet learning continues effectively, AI-driven shifts in educational methods need not impair intellectual growth. Active learning strategies like project-based and inquiry-based methods consistently show effectiveness in developing critical thinking and synthesis skills.
Instead of fearing AI, education should embrace diverse learning strategies, ensuring students engage deeply and meaningfully, whatever tools they use.
Thanks for your comment. While I agree that writing papers is far from the only way to develop critical thinking skills (though I don't think the other ways are direct substitutes for putting thoughts into words), I have to say that I think your typewriter analogy is specious. Yes, writing by hand is different from writing with a typewriter, but neither a pencil nor a typewriter will do your writing for you. That's a change of an entirely different order. Muddying the waters does not clear the waters.
Fair point, though perhaps we’re both muddying waters worth exploring. Just as the typewriter didn’t ‘do the thinking,’ AI also doesn’t inherently replace thoughtful engagement. It’s merely easier to misuse. The pencil and typewriter didn’t guarantee better writing any more than AI guarantees worse learning; each shift simply asks us to rethink how best to inspire genuine thought. Perhaps the real skill for educators and learners alike is adapting wisely to our tools, AI included.
Urgh I am trying to hold back but every sentence then in your follow up was also nonsensical. Worst is the “maybe the real skill is that we adapt!” pablum.
You’re right that I was too cryptic earlier. That’s on me. Not great modeling for an educator, so let me try again more clearly.
Socrates, through Plato, famously said that writing was a terrible invention because it would make students rely on the written word instead of memory. People also feared that printing presses, typewriters, and even calculators would ruin learning. But again and again, we adapted. And while AI is definitely a bigger shift, it’s still part of this long pattern: the tool changes, people panic, and eventually, education evolves.
My point about typewriters wasn’t that they are the same as AI. Of course they’re not! But both are tools, and tools don’t replace knowledge or thought unless we let them. A typewriter didn’t make someone a better thinker, and neither does AI. What matters is the *learning environment*: what we ask students to do with the tools we give them.
Many educators are already past the “sky is falling” phase. Like teachers who once stopped grading for penmanship and started assessing reasoning or insight, we now need to find new ways to assess critical thinking that don’t rely solely on polished essays. There’s real, hard work ahead, and I think that’s the direction we need to head.
Thanks for bearing with me, and I hope this version is clearer.
Can't agree with your substitution here. As Mr. Carr already pointed out, the typewriter does not do the writing for someone who has shifted over from handwriting. As I honed my writing skill over a long period, it was putting together sentences and paragraphs that cohere and mean something that mattered, not whether I did it pen-on-paper or using a typewriter to prepare a suitable final version for submission. Later adoption of the wordprocessor followed the first path mentioned above, namely, it enhanced by making more efficient and flexible a skill I had already developed. So far as I can tell, occasional typos notwithstanding, keyboards have not eroded my writing skill. However, I can attest that when I do sit down to write on paper, the flow of ideas is subtly different.
Totally agree that typing didn’t erode your writing skill, because it followed the development of that skill. That’s the key. My point is that no tool (pen, typewriter, or AI) guarantees thought; the learning challenge is to design environments where thinking still has to happen. I share more on how we might do that forward-facing, here: https://substack.com/@glagna/note/p-164577138?r=5rlap6&utm_medium=ios&utm_source=notes-share-action
"What we should be more concerned about is how AI cheats students." Fully agree with your emphasis on what students are losing in the break-neck adoption of AI tools. My husband and I recently attended a talk at the Perimeter Institute on AI and education, and it was the most hopeful perspective that I have come across: https://schooloftheunconformed.substack.com/p/learning-fast-and-slow-why-ai-will
Thanks for your writing!
Well done! I hope no AI was hurt in the production of this post.
Learning comes from struggle. It's frustrating, time-consuming, but when the light goes on, you remember it because you've had to work to flip that switch of discovery.
When "learning" is too easy, it ceases to be memorable. It becomes ephemeral and forgettable.
I'm not even sure students are swimming in the shallows here--more like they're dipping toes in the ocean and then moving on, never immersing themselves, never feeling the thrill of waves and currents of knowledge, never getting tossed around but emerging smarter and wiser from the experience.
When I graduated college in 2019, I never would have guessed it was the final year of a pure college experience. Once the last Covid restrictions subsided in 2023, generative AI was already off to the races. It’s all very sad.
Beautifully explained with a great comparison to the Industrial Revolution. It’s wild how the IR also revolutionized not just work but society so I wonder what the AI revolution will do to society.
Nice post! I just wrote about AI in college classrooms on my substack, here:
https://profbachman.substack.com/p/will-we-still-have-professors
Some of my readers recommended your post as a follow-up. We have many of the same points, but I make a distinction between two essentially different roles of education: mastering skills (e.g. writing, problem solving, etc), where I agree AI can be detrimental for the reasons you have so eloquently stated, and understanding concepts (e.g. what is the quantum mechanical model of the atom?), where AI is a fantastic tool. That balance between skill acquisition (where AI bad) and concept mastery (where AI good) makes the whole AI-in-schools thing very tricky!
Thank you for another interesting post.
I would definitely agree with footnote 2, which may be pushed further in the context of tertiary study when it becomes simply about needing that 'piece of paper' in order to get the job.
At its most naked, this can be seen, for example, in Australia where degrees are more or less sold to foreign students who may or may not even have proficiency in English. What does the degree mean in such a situation? Some sort of entry token to particular types of job, perhaps.
'Education' is one of the nation's top exports, and this practice began long before A.I. entered the equation. In such a mockery of learning, the use of A.I. to write papers is simply the next logical step.
Excellent essay, thanks Nick.
I've found Punya Mishra's 2x2 on AI and domain expertise to be a helpful guide in thinking about the difference between how experts with existing domain knowledge can interact with and leverage AI vs. how students without the hard-won expertise are tricked into thinking they have it without building it. https://punyamishra.com/2025/02/13/the-genai-and-expertise-paradox-why-it-makes-expert-work-more-important-but-harder/
I feel strongly that the best antidote to all of this is going to require educators to double down on building communities of trust with their students. The temptation of product without process is so strong for our busy and overwhelmed students who are continuously hearing that they need to keep up with the cutting edge or be left behind. The best thing we can do is help to articulate the false promises and ask them to trust us in helping them navigate through the storm. At some point I do think we'll need to ask them to tie themselves to the mast of the ship to endure the journey through the cacophony of the AI sirens' calls.
"The fact is the sweetest dream that labor knows"
But the fact is not the product of labor but labor itself as an action
I fully agree with your passage about the "paper" being only a proxy of understanding and that the real gain is in the journey not really in the destination.
That has been my most contentious point regarding all the AI hype.
Regarding education, I think the question goes deeper than AI discouraging learning. As I see it, the biggest crisis in education right now is that it cannot state a coherent narrative of what Education is for
This is a question that all cultures must answer, preferably consciously.
Today, the answer is often that education is an economic leverage, meaning being a good consumer. Or as you pointed, a customer
This is the most important Substack post I’ve read in weeks. Thank you for laying out so clearly the skill paths that people go down when they adopt a new technology. Your powers of explanation are on full display here. Bravo!
PS One of my concerns is that people may acknowledge that students are learning less but they will claim that learning is no longer as relevant as it once was because they will point to how students are (seemingly) solving so many new problems and advancing humanity’s knowledge with the help of AI. I witnessed the same thing with digital humanities projects: lots of cool ways to display information using graphics and webpages, but it was never clear if anyone was learning much that was new or important.
I find stark parallels between education (teaching) and the trucking industry (the largest first time employer in the USA). Technology doesn't at first replace workers, rather it competes with them. It creates lower wages, higher turnover, precarity and de-regulation. As workers become transcient laborers, it then becomes much easier to replace them, so disempowered they've become. We see this in education. No lack of teachers, just lack of teachers willing to do the job for peanuts, no respect and the stress/work to having a life ratio. The real issue isn't replacing human labor - it is a social fabric that doesn't distribute the gains to the working class. AI should allow us to pay teachers more, free up for more time on real learning/interaction but rather the opposite occurs because we focus on economic value, not social value.
Spot on.
I believe ne of the biggest failures in our schools during this technological age is that we handed out tablets where students learned only to swipe and tap. We neglected to teach computer skills like typing & using basic software. Imagine AI is available to you and you don't know how to properly type. I wrote more about that here:
https://open.substack.com/pub/dearchristianparent/p/screens-in-schools-digital-preparation?r=3huc9s&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false
Could the answer be more exam conditions essay writing? Ie you are put in library for 6 hours, phones put in a box and told to write an essay.
That would test whether these students have actually acquired the skills?
“Glass Cage” is a great work and this use of AI is the latest version of deskilling.
Thank you.
Your excellent article moved me to restack it several times with the perspective that the big tech oligarchs are drug pushers. I hope that "junkie" perspective adds value to this whole, critical debate.
Ps I'm new here on substack so it will be much appreciated if you can also support me e.g. by restacking my restacks to your audience...
Ever since 2021 when I read The Shallows, I‘ve thought of meaninglessness as the biggest threat when it comes to the development of AI.
Here’s a Swedish poem called “In motion” by Karin Boye (1927) to be used in your next book!
Den mätta dagen, den är aldrig störst.
Den bästa dagen är en dag av törst.
Nog finns det mål och mening i vår färd -
men det är vägen, som är mödan värd.
Det bästa målet är en nattlång rast,
där elden tänds och brödet bryts i hast.
På ställen, där man sover blott en gång,
blir sömnen trygg och drömmen full av sång.
Bryt upp, bryt upp! Den nya dagen gryr.
Oändligt är vårt stora äventyr.
Do you want to translate it, or shall I use ChatGPT?
Ha! Let’s leave it to a human being (David McDuff). Here goes:
The sated day is never first.
The best day is a day of thirst.
Yes, there is goal and meaning in our path -
but it's the way that is the labour's worth.
The best goal is a night-long rest,
fire lit, and bread broken in haste.
In places where one sleeps but once,
sleep is secure, dreams full of songs.
Strike camp, strike camp! The new day shows its light.
Our great adventure has no end in sight.
Interesting article. I admit it raises valid concerns about AI potentially diminishing learning through automation. However, I can't help noticing that it mistakenly assumes essay writing is essential for developing critical thinking, overlooking numerous other effective learning pathways. To illustrate this point, let me rephrase the tenth paragraph as it would have sounded a few decades ago, when similar fears arose when typewriters replaced handwriting:
“Typewriters enable students to produce text without doing the careful work of handwriting. Rather than meticulously crafting each letter by hand, students press keys effortlessly. Rather than carefully constructing elegant calligraphy strokes that demonstrate discipline and attention to detail, they now produce text mechanically. And rather than expressing (and refining) their thoughts through the deliberate, artistic act of penmanship, they quickly type out a first draft or even a final one.”
Just as handwriting is no longer central yet learning continues effectively, AI-driven shifts in educational methods need not impair intellectual growth. Active learning strategies like project-based and inquiry-based methods consistently show effectiveness in developing critical thinking and synthesis skills.
Instead of fearing AI, education should embrace diverse learning strategies, ensuring students engage deeply and meaningfully, whatever tools they use.
Thanks for your comment. While I agree that writing papers is far from the only way to develop critical thinking skills (though I don't think the other ways are direct substitutes for putting thoughts into words), I have to say that I think your typewriter analogy is specious. Yes, writing by hand is different from writing with a typewriter, but neither a pencil nor a typewriter will do your writing for you. That's a change of an entirely different order. Muddying the waters does not clear the waters.
Fair point, though perhaps we’re both muddying waters worth exploring. Just as the typewriter didn’t ‘do the thinking,’ AI also doesn’t inherently replace thoughtful engagement. It’s merely easier to misuse. The pencil and typewriter didn’t guarantee better writing any more than AI guarantees worse learning; each shift simply asks us to rethink how best to inspire genuine thought. Perhaps the real skill for educators and learners alike is adapting wisely to our tools, AI included.
Your analogy was terrible.
Thanks for your insight!
Urgh I am trying to hold back but every sentence then in your follow up was also nonsensical. Worst is the “maybe the real skill is that we adapt!” pablum.
You’re right that I was too cryptic earlier. That’s on me. Not great modeling for an educator, so let me try again more clearly.
Socrates, through Plato, famously said that writing was a terrible invention because it would make students rely on the written word instead of memory. People also feared that printing presses, typewriters, and even calculators would ruin learning. But again and again, we adapted. And while AI is definitely a bigger shift, it’s still part of this long pattern: the tool changes, people panic, and eventually, education evolves.
My point about typewriters wasn’t that they are the same as AI. Of course they’re not! But both are tools, and tools don’t replace knowledge or thought unless we let them. A typewriter didn’t make someone a better thinker, and neither does AI. What matters is the *learning environment*: what we ask students to do with the tools we give them.
Many educators are already past the “sky is falling” phase. Like teachers who once stopped grading for penmanship and started assessing reasoning or insight, we now need to find new ways to assess critical thinking that don’t rely solely on polished essays. There’s real, hard work ahead, and I think that’s the direction we need to head.
Thanks for bearing with me, and I hope this version is clearer.
Can't agree with your substitution here. As Mr. Carr already pointed out, the typewriter does not do the writing for someone who has shifted over from handwriting. As I honed my writing skill over a long period, it was putting together sentences and paragraphs that cohere and mean something that mattered, not whether I did it pen-on-paper or using a typewriter to prepare a suitable final version for submission. Later adoption of the wordprocessor followed the first path mentioned above, namely, it enhanced by making more efficient and flexible a skill I had already developed. So far as I can tell, occasional typos notwithstanding, keyboards have not eroded my writing skill. However, I can attest that when I do sit down to write on paper, the flow of ideas is subtly different.
Totally agree that typing didn’t erode your writing skill, because it followed the development of that skill. That’s the key. My point is that no tool (pen, typewriter, or AI) guarantees thought; the learning challenge is to design environments where thinking still has to happen. I share more on how we might do that forward-facing, here: https://substack.com/@glagna/note/p-164577138?r=5rlap6&utm_medium=ios&utm_source=notes-share-action