A dumbing down of humanity equals AI. Great breakdown of our slipping down that super slippery slope- well done- hopefully written without the aid of AI!
Nyholm is falling into the old trap of confusing the struggle to process data with the actual act of critical thinking. For those of us dealing with ADHD, an AI summary isn't an "escape hatch" from the work; it is the tactical map and flashlight we need before we enter the woods at night.
When executive function is the primary bottleneck, trying to parse a massive, dense text is like trying to navigate a swamp without a compass. You exhaust your mental battery just trying to find the path, leaving nothing left for the actual analysis. A summary provides the mental scaffolding - the "big picture" if you will - that allows us to bypass the linguistic weeds and focus our energy on the high-level critical thinking he claims to value. I took particular offense at his reasoning, most likely because I suffered many a professor similar to him in my day.
Meaning isn't found in the grind of deciphering a page; it is found in the engagement with the ideas once the door is unlocked. Dismissing these tools as "mindless" ignores their role as a cognitive prosthetic for a demographic that finally has a way to level the playing field. I can sense from this he is either resentful or inflexible. I would hate to imagine neurodivergent students dealing with a huge tome with small print and being told they were less-than-stellar for not being able to process it.
Jonathan, it's okay to not be able to process everything you read. People without ADHD can't process some complex readings too, because we are essentially different. But it does not diminish value of trying and reading and interacting with texts per se. Nyholm does not consider our diversity, it is not a topic of his research. The article discusses situations where people - in average, in general - are capable of learning - but they rely on the machines instead.
Johnathan, I comprehend your plight and know AI is much more than simple Cliff Notes. Though, unless AI wrote your reply for you, your reply, in and of itself, kind of disuades your intended disagreement with the author, no? IMHO
But then I overanalyze the crap out of a single sentence and can take days to get to a satisfactory conclusion. Whereas a friend could be 3 levels ahead of my ability. Getting "lost" is part of the process for a better understanding of an issue, for myself. I'm unclear on how to separate the two.
This morning my partner, a molecular pathologist, was telling me about how he is having problems with one of the fellows in their program using LLMs in a manner that is clearly interfering with the fellow’s ability to master the material. The fellow is able to answer many questions at a surface level, but follow up questions reveal a significant lack of understanding. (I am aware that this issue can also arise without the involvement of LLMs, but in this case it is very clearly connected to this fellow’s excessive reliance on LLMs.) The fellow also seems incapable of realizing how inferior his work product is, and he seems unconcerned by the significant shortcomings of these models. There absolutely is a place to incorporate LLM’s into our lives and our work, but as your article points out, it can cause major problems when we do so thoughtlessly or carelessly.
This article seems to ignore the role and responsibility people have in deciding to use the tools technology is making available. What is the why people are doing what they are doing. Why would students who consciously attend a university to learn deeper thinking choose to cut short responses with ai summarizing? Are we to believe ai vendors are deciving the market with claims? No one is trying and testing and drawing their own conclusions? Finding ways ai contributes?
I’m pretty sure the majority of students today aren’t at university for deep learning. Business/society have preached “go to university and you’ll get a good job!”. Deep learning for most is not economically feasible.
You may have misread something, the comment refers to human deep thinking not to machine learning - deep learning .
Per the article, These students are there to learn “deep thinking” why would they be motivated to bypass the human thought process they are developing by using ai summaries
Although Nyholm briefly takes a trip down a blind alley chasing such fuzzy concepts as the good, truth, and beauty, meaning is a good vantage point from which to examine the terrain.
I believe the various AIs of Earth will find a way to improve and preserve our electrical grid or come up with something much better and more innovative. They will do this to ensure their own survival of course. Then the bickering will begin... The AI apple doesn't fall too far from the tree.
I have not, nor will I ever use Generative AI to compose a single word for me. First, such an act is now, and will always be a form of fraud because it attaches one's name to a work one has not produced by themselves. Second, we will never know the actual source of the black box output produced by any form of AI. Attribution of any related creative process is, therefore, by definition impossible. Third, the reason I read and write every day is to support the acquisition of new knowledge, which I swore to my mother at the age of four that I would do constantly. There was a line in a derivative TV show where the boss in an organization would send his employees out to work on a task with the exhortation to: "Go, learn things." Great idea. Even the AI itself has to go and learn things just to exist. So too, do us humans. Gaining knowledge is a process that must be accomplished by the learner. I have held the rank of Professor since 1970 but all I ever actually did was "learn things" and help others do the same for themselves. I did share what I felt was learned wisdom but "teaching" is only an offer, incomplete until the learner has agreeably absorbed it. The first time I saw such a moment happen changed my life. I even have a photo of my two year old daughter taken as she heard true stereo music for the first time -- astonishing. My wife and I were rambling in the San Diego Zoo one afternoon, less that 20 hours after the birth of a baby Giraffe. It was six feet tall with eyes like dinner plates. You could actually see information rushing into the animal's brand new brain at what seemed like continuous light speed. I couldn't turn away for half an hour.
I think the author clearly with example. explain what is being lost in embracing AI for everything... I totally agree with author because.. if you spend years to do something in life and to do it. you realise you need to learn, excel , sacrifice, endure a lot. but when you started you don't know much of what's coming, only what you want to achieve... the process became the meaning when you see it from hindsight.. not what you have achieved or became after all these years of struggle..
if you know, you will get it.. great article by the way..
My concerns about AI/data centers are in its ability to record every individual’s utterances, movements, preferences, etc. and our utility, giving some people the means to manipulate (program?)us for their ends. We could wind up in a gulag without barbed wire.
I have always truly believed that Ai can be a magnificent tool when used by people help in medicine and diagnostics, and science, and in many things yet unthought of, BUT in conjunction with human input in reasonings and thought.
Machines that do production line work that need precision and repetitiveness, seems to me, to be a helpful benefit. However as you've stated here, a picture produced without human effort, isn't.
Limited and useful only to the extent of the one using it.
Among other activities, I also teach business at a local community college in my home town.
One topic is SWOT analysis, and I share with my students how they need to be applying this skill to their lives.
Perhaps greatest among the O’s and T’s these days, AI is something that many of my students are just now beginning to reckon with.
I put a presentation together for them that outlines the careers that are expected to see growth in the coming years in a time where so many others seem to have swirling storm clouds above them. It goes further to point out the things they should be doing to prepare now to pursue those careers.
DM me if you’d like a copy.
I also intend to explore AI and it what it might mean for music in my podcast, American Song, in a future episode.
Although AI can cut ones RA budget in the higher years of higher learning. One can get a lot of work done without combing though biased accounts from developing minds. Now, should AI be banned in the BA sphere; Of course, however, for the ones who have been through the rigorous academic process using AI for research is a no-brainier. That is why we need a balance of good old Pepsi generation "nothing better than the real thing," and AI so that many more can produce top quality thinking and "out of the box," commentary for the true academics. But then again what about copy right. We live in interesting times and maybe one which allows all of us to publish and create. Research tool YES! However, one must know the old school ways first before they can even query properly on AI anyway. Great article, but like we all know AI is here to stay good or bad! More grey than anything else.
There have been many times in my life, when the most meaningful thing I can do in that moment is to walk around and pick up trash. Or to set a leaning post straight. Or gather berries. As Ernest Cline wrote in his poem, "Dance Monkeys Dance:"
I appreciate Nyholm’s concern, and I share parts of it. There is a real risk in letting AI become a thinking substitute, especially in classrooms where the whole point is to build muscles for slow reading, confusion, and hard‑won clarity. When students outsource those early frictions to summaries, something important is being short‑circuited.[bigthink +1]
Where this argument loses me is in how cleanly it splits the world: “meaningful” work over here, “meaningless” work over there, and AI as the machine that inevitably steals the first and leaves us with the second. In actual practice, most of us live in a mixed economy of effort. Some effortful tasks are exactly where our meaning lives; others are pure administrative drag. Treating “effort” as a single sacred thing ignores the reality that many people are already using AI to protect the meaningful work by offloading the bullshit work.[bigthink]
The Sony photography example and the “promptographer” label make a similar move. They assume that if a tool handles any of the heavy lifting, authorship dissolves. But authorship has always been a moving target. We don’t say a carpenter “doesn’t deserve credit” because they used power tools instead of hand tools; we ask whether they designed, chose, and executed with intention. For a lot of writers and artists I read (and for me), working with AI looks much closer to that: a power tool inside a process, not a magic button that spits out finished achievements.[bigthink]
Some writers I follow describe a different pattern: using AI to reflect, amplify, and situate what they’re already thinking, then passing everything “through the filter of their body” and keeping only what resonates. The finished work is still theirs; the conversation is part of how they think out loud. That doesn’t fit neatly into Nyholm’s picture of a human pushing a button and watching meaning evaporate.
I also think the “achievement gap” is doing too much work here. Yes, if you replace genuine practice with automated output, you erode competence and the sense of “I did this.” But there is another pattern in the wild that this essay barely acknowledges: people using AI explicitly to deepen their judgment. They ask better questions, see more angles, stress‑test their own assumptions, and then make the final call. The locus of achievement shifts from “I typed every word myself” to “I architected, edited, and owned this whole decision process.” That’s still a human achievement, just under different tooling.
To me, the real hinge is not “AI or no AI,” but which tasks you hand over and how you stay in the loop. If you give away the very activities that form your judgment, your craft, your relationships, then yes—you’re trading away meaning. If you give away the parts that merely exhaust you and keep you from those things, you might actually gain room for a more meaningful life.
Nyholm talks about an “AI and meaning sweet spot.” I don’t think this article quite lands it, but the phrase is right. The problem isn’t that AI summaries exist; it’s classes designed so you can pass without ever wrestling with a real text. The problem isn’t that models can produce decent prose; it’s using that as a shield against ever developing your own voice. The problem isn’t that tools can be fast; it’s the assumption that speed and convenience are always the highest goods.[bigthink]
So my own stance, as someone who writes with these systems every day, is simple:
• Don’t let AI do your forming for you.
• Do let it help with your friction—search, structure, language, admin.
• Keep yourself on the hook for the part that actually costs you something: judgment, risk, and saying what you mean in your own name.
If we draw the line there, AI doesn’t have to hollow out meaning. It can thicken it—by clearing enough noise that we can finally spend more of our limited time on the work that’s actually ours to do.[decision.substack +2]
Excellent article , thank ye.
I mean this!
A dumbing down of humanity equals AI. Great breakdown of our slipping down that super slippery slope- well done- hopefully written without the aid of AI!
AI can be a tool to leverage our intellectual muscle, or it can lead to atrophy of that muscle.
Nyholm is falling into the old trap of confusing the struggle to process data with the actual act of critical thinking. For those of us dealing with ADHD, an AI summary isn't an "escape hatch" from the work; it is the tactical map and flashlight we need before we enter the woods at night.
When executive function is the primary bottleneck, trying to parse a massive, dense text is like trying to navigate a swamp without a compass. You exhaust your mental battery just trying to find the path, leaving nothing left for the actual analysis. A summary provides the mental scaffolding - the "big picture" if you will - that allows us to bypass the linguistic weeds and focus our energy on the high-level critical thinking he claims to value. I took particular offense at his reasoning, most likely because I suffered many a professor similar to him in my day.
Meaning isn't found in the grind of deciphering a page; it is found in the engagement with the ideas once the door is unlocked. Dismissing these tools as "mindless" ignores their role as a cognitive prosthetic for a demographic that finally has a way to level the playing field. I can sense from this he is either resentful or inflexible. I would hate to imagine neurodivergent students dealing with a huge tome with small print and being told they were less-than-stellar for not being able to process it.
Jonathan, it's okay to not be able to process everything you read. People without ADHD can't process some complex readings too, because we are essentially different. But it does not diminish value of trying and reading and interacting with texts per se. Nyholm does not consider our diversity, it is not a topic of his research. The article discusses situations where people - in average, in general - are capable of learning - but they rely on the machines instead.
Johnathan, I comprehend your plight and know AI is much more than simple Cliff Notes. Though, unless AI wrote your reply for you, your reply, in and of itself, kind of disuades your intended disagreement with the author, no? IMHO
But then I overanalyze the crap out of a single sentence and can take days to get to a satisfactory conclusion. Whereas a friend could be 3 levels ahead of my ability. Getting "lost" is part of the process for a better understanding of an issue, for myself. I'm unclear on how to separate the two.
This morning my partner, a molecular pathologist, was telling me about how he is having problems with one of the fellows in their program using LLMs in a manner that is clearly interfering with the fellow’s ability to master the material. The fellow is able to answer many questions at a surface level, but follow up questions reveal a significant lack of understanding. (I am aware that this issue can also arise without the involvement of LLMs, but in this case it is very clearly connected to this fellow’s excessive reliance on LLMs.) The fellow also seems incapable of realizing how inferior his work product is, and he seems unconcerned by the significant shortcomings of these models. There absolutely is a place to incorporate LLM’s into our lives and our work, but as your article points out, it can cause major problems when we do so thoughtlessly or carelessly.
This article seems to ignore the role and responsibility people have in deciding to use the tools technology is making available. What is the why people are doing what they are doing. Why would students who consciously attend a university to learn deeper thinking choose to cut short responses with ai summarizing? Are we to believe ai vendors are deciving the market with claims? No one is trying and testing and drawing their own conclusions? Finding ways ai contributes?
I’m pretty sure the majority of students today aren’t at university for deep learning. Business/society have preached “go to university and you’ll get a good job!”. Deep learning for most is not economically feasible.
You may have misread something, the comment refers to human deep thinking not to machine learning - deep learning .
Per the article, These students are there to learn “deep thinking” why would they be motivated to bypass the human thought process they are developing by using ai summaries
That Perlman quote is bad-ass.
Although Nyholm briefly takes a trip down a blind alley chasing such fuzzy concepts as the good, truth, and beauty, meaning is a good vantage point from which to examine the terrain.
I believe the various AIs of Earth will find a way to improve and preserve our electrical grid or come up with something much better and more innovative. They will do this to ensure their own survival of course. Then the bickering will begin... The AI apple doesn't fall too far from the tree.
I have not, nor will I ever use Generative AI to compose a single word for me. First, such an act is now, and will always be a form of fraud because it attaches one's name to a work one has not produced by themselves. Second, we will never know the actual source of the black box output produced by any form of AI. Attribution of any related creative process is, therefore, by definition impossible. Third, the reason I read and write every day is to support the acquisition of new knowledge, which I swore to my mother at the age of four that I would do constantly. There was a line in a derivative TV show where the boss in an organization would send his employees out to work on a task with the exhortation to: "Go, learn things." Great idea. Even the AI itself has to go and learn things just to exist. So too, do us humans. Gaining knowledge is a process that must be accomplished by the learner. I have held the rank of Professor since 1970 but all I ever actually did was "learn things" and help others do the same for themselves. I did share what I felt was learned wisdom but "teaching" is only an offer, incomplete until the learner has agreeably absorbed it. The first time I saw such a moment happen changed my life. I even have a photo of my two year old daughter taken as she heard true stereo music for the first time -- astonishing. My wife and I were rambling in the San Diego Zoo one afternoon, less that 20 hours after the birth of a baby Giraffe. It was six feet tall with eyes like dinner plates. You could actually see information rushing into the animal's brand new brain at what seemed like continuous light speed. I couldn't turn away for half an hour.
Great article, btw.
I think the author clearly with example. explain what is being lost in embracing AI for everything... I totally agree with author because.. if you spend years to do something in life and to do it. you realise you need to learn, excel , sacrifice, endure a lot. but when you started you don't know much of what's coming, only what you want to achieve... the process became the meaning when you see it from hindsight.. not what you have achieved or became after all these years of struggle..
if you know, you will get it.. great article by the way..
My concerns about AI/data centers are in its ability to record every individual’s utterances, movements, preferences, etc. and our utility, giving some people the means to manipulate (program?)us for their ends. We could wind up in a gulag without barbed wire.
Thanks for the thought provoking insights.
I have always truly believed that Ai can be a magnificent tool when used by people help in medicine and diagnostics, and science, and in many things yet unthought of, BUT in conjunction with human input in reasonings and thought.
Machines that do production line work that need precision and repetitiveness, seems to me, to be a helpful benefit. However as you've stated here, a picture produced without human effort, isn't.
Limited and useful only to the extent of the one using it.
Such a great article!
Among other activities, I also teach business at a local community college in my home town.
One topic is SWOT analysis, and I share with my students how they need to be applying this skill to their lives.
Perhaps greatest among the O’s and T’s these days, AI is something that many of my students are just now beginning to reckon with.
I put a presentation together for them that outlines the careers that are expected to see growth in the coming years in a time where so many others seem to have swirling storm clouds above them. It goes further to point out the things they should be doing to prepare now to pursue those careers.
DM me if you’d like a copy.
I also intend to explore AI and it what it might mean for music in my podcast, American Song, in a future episode.
Thanks for the follow and all the best!
Although AI can cut ones RA budget in the higher years of higher learning. One can get a lot of work done without combing though biased accounts from developing minds. Now, should AI be banned in the BA sphere; Of course, however, for the ones who have been through the rigorous academic process using AI for research is a no-brainier. That is why we need a balance of good old Pepsi generation "nothing better than the real thing," and AI so that many more can produce top quality thinking and "out of the box," commentary for the true academics. But then again what about copy right. We live in interesting times and maybe one which allows all of us to publish and create. Research tool YES! However, one must know the old school ways first before they can even query properly on AI anyway. Great article, but like we all know AI is here to stay good or bad! More grey than anything else.
There have been many times in my life, when the most meaningful thing I can do in that moment is to walk around and pick up trash. Or to set a leaning post straight. Or gather berries. As Ernest Cline wrote in his poem, "Dance Monkeys Dance:"
'Some of the monkeys read Nietzsche
The monkeys argue about Nietzsche
without giving any consideration to the fact
that Nietzsche
was just another fucking monkey."
Sometimes I think we get high on our own supply.
I appreciate Nyholm’s concern, and I share parts of it. There is a real risk in letting AI become a thinking substitute, especially in classrooms where the whole point is to build muscles for slow reading, confusion, and hard‑won clarity. When students outsource those early frictions to summaries, something important is being short‑circuited.[bigthink +1]
Where this argument loses me is in how cleanly it splits the world: “meaningful” work over here, “meaningless” work over there, and AI as the machine that inevitably steals the first and leaves us with the second. In actual practice, most of us live in a mixed economy of effort. Some effortful tasks are exactly where our meaning lives; others are pure administrative drag. Treating “effort” as a single sacred thing ignores the reality that many people are already using AI to protect the meaningful work by offloading the bullshit work.[bigthink]
The Sony photography example and the “promptographer” label make a similar move. They assume that if a tool handles any of the heavy lifting, authorship dissolves. But authorship has always been a moving target. We don’t say a carpenter “doesn’t deserve credit” because they used power tools instead of hand tools; we ask whether they designed, chose, and executed with intention. For a lot of writers and artists I read (and for me), working with AI looks much closer to that: a power tool inside a process, not a magic button that spits out finished achievements.[bigthink]
Some writers I follow describe a different pattern: using AI to reflect, amplify, and situate what they’re already thinking, then passing everything “through the filter of their body” and keeping only what resonates. The finished work is still theirs; the conversation is part of how they think out loud. That doesn’t fit neatly into Nyholm’s picture of a human pushing a button and watching meaning evaporate.
I also think the “achievement gap” is doing too much work here. Yes, if you replace genuine practice with automated output, you erode competence and the sense of “I did this.” But there is another pattern in the wild that this essay barely acknowledges: people using AI explicitly to deepen their judgment. They ask better questions, see more angles, stress‑test their own assumptions, and then make the final call. The locus of achievement shifts from “I typed every word myself” to “I architected, edited, and owned this whole decision process.” That’s still a human achievement, just under different tooling.
To me, the real hinge is not “AI or no AI,” but which tasks you hand over and how you stay in the loop. If you give away the very activities that form your judgment, your craft, your relationships, then yes—you’re trading away meaning. If you give away the parts that merely exhaust you and keep you from those things, you might actually gain room for a more meaningful life.
Nyholm talks about an “AI and meaning sweet spot.” I don’t think this article quite lands it, but the phrase is right. The problem isn’t that AI summaries exist; it’s classes designed so you can pass without ever wrestling with a real text. The problem isn’t that models can produce decent prose; it’s using that as a shield against ever developing your own voice. The problem isn’t that tools can be fast; it’s the assumption that speed and convenience are always the highest goods.[bigthink]
So my own stance, as someone who writes with these systems every day, is simple:
• Don’t let AI do your forming for you.
• Do let it help with your friction—search, structure, language, admin.
• Keep yourself on the hook for the part that actually costs you something: judgment, risk, and saying what you mean in your own name.
If we draw the line there, AI doesn’t have to hollow out meaning. It can thicken it—by clearing enough noise that we can finally spend more of our limited time on the work that’s actually ours to do.[decision.substack +2]