Then is your version of right hemisphere attention to go with one's gut?
Do you actually consider statistics from a comprehensive perspective?
We have seen the triumph of left-hemisphere obscene wealth around the planet over the past 40 years, and you want to say that women are unhappier because now they don't have to apply for work through separate employment ads? That somehow my wife has remained deeply happy despite the fact that the tireless work of feminists made it infinitely easier for her to go to medical school?
Nobody is attacking you for your capacity to integrate reasonable analysis and intuition. They are pointing out that you have let some deep misreadings of statistics enable you to bypass both hemispheres in favor of, well, I think you can intuit what I mean.
If you think duality is the opposite of non duality, there's the whole problem in a nutshell. Pure left brain thinking! Maybe several thousand years of contemplative wisdom is not irrelevant in the face of the fact that, as you said to me some time back, meditative practice of any kind is a left brain mistake.
Thank you for these reflections. The discussion of the 'woke' impulse reminded me of a quote from Rudolf Steiner:
“Take the following case—It may happen in life that two people live together for a long time and that through the strange forces playing out of the unknown regions of the astral [soul] body and Ego of the one person into the astral body and Ego of the other (these forces remain in the hidden regions), the one has in relation to the other a real desire for torment, a kind of need for cruelty. It may be that the one person who has this desire for torment, this need for cruelty, has no inkling whatever of the existence of these emotions in the astral body and Ego; he may build up about the things he does out of this urge to cruelty, a whole number of ideas which explain the actions on quite other grounds. Such a person may tell us that he has done this or that to the other person for one reason or another; these reasons may be very clever and yet they do not express the truth at all. For in ordinary life, what we all-too-often picture as the motives of our own actions, indeed of our own feelings, frequently stands, as I say, in a very, very distant connection with what is really living and weaving in our inner being. It may be that the Luciferic power is actually preventing the person concerned from realising the nature of this urge for cruelty, of these impulses to do all kinds of things to the other person, and that under the influence of this Luciferic power everything he says about the reasons merely spreads a cover over what is actually present in the soul. The reasons we devise in our consciousness may often be cut out for hiding from us, disguising what is actually living and working in the soul. These reasons are too often of a character which indicates a desire for self-justification, for we should find ourselves just as antipathetic as the professor of philosophy of whom I told you. We should not at all like what is in our soul if we had to acknowledge what kind of instincts and emotions are really holding sway. And because we have to protect ourselves from the sight of our own soul-being, we discover, with the help of these reasons, all kinds of things that guarantee us protection, because they deceive us about what is actually the ruling force in the soul. Just as it is true that the external world becomes a Maya to us because of the peculiar character of our faculty to form mental pictures, it is also true that what we have to say about ourselves in ordinary life is, to a very, very great extent, Maya.” (GA 161, I)
Forgive me if I have missed something here. Is Steiner saying that there are forces at work in our souls of which we are ignorant and over which we have no control? I'm sure he has an antidote to this.
Yes, for sure. His entire life's work could be considered the antidote. It goes in the direction of developing imagination and intuition as consciously directed cognitive faculties, which help us unveil the deeper soul forces that we are normally ignorant of and modulate them, such that they begin to coincide with our higher human ideals.
This path is laid out in a philosophical form in his early work, The Philosophy of Freedom (or "Intuitive Thinking as a Spiritual Path").
Agreed, many keen public thinkers these days seem to be instinctively probing the etheric spectrum of consciousness and therefore gaining insights into the second-order processes that make our intelligent navigation of life possible. With an introduction to Steiner, they would simply be making what they are already doing with their intuitive activity more and more conscious.
I am a great admirer of your work, and I share your antipathy to coercions of all kinds, left liberal that I am. I just posted this this morning, since it's part of my work against the darkness I thought it might brighten things a bit. https://billionworlds.substack.com/p/escucha-la-luz-en-la-noche-listen
Seems the universe has a voice that those tuned to it can decipher at least enough to jiggle the brighter minds to acknowledge canaries, actively pursue creative paths that strain the analog mind and with hope be the tortoise that wins in the end. 🕊️
I was about to say this thread could use some poetry when I came to your post, Elizabeth, albeit disguised as prose. For Sir Iain's work must be read as poetry, albeit also disguised as prose.
Indeed, risky. Though I might plead that I am from the other side of the pond, and "Sir" is respectful all round the world without knighthood being required. That being said, Iain might yet be knighted. I think he will be by posterity, as one of the great scientist-writer-philosophers of our time.
Speaking on "cognitive freedom" at Duke and concerned about wokeness and the sleep of reason? You might find the case of Paul Griffiths, ex-Duke divinity professor, relevant: "Duke divinity professor calls diversity training ‘a waste,’ faces discipline" at https://www.newsobserver.com/news/local/education/article149462029.html
Yes, imagine that. Nearly a billion people around the planet have explored various diversity training initiatives, and between you and Iain, I'll bet you can easily come up with 50 or more horrifying examples.
Now let's compare with wellness examples about LH being logical and RH being emotional.
Would you reconsider the value of anecdotal examples if we came up with 1000 examples of horrifying DEI initiates and 7 million about the hemispheres?
No doubt he does. But would you say the same if he were friends with Putin? It’s just a really base and pathetic response I would expect from a teenager. RD is pals with JD Vance who is perhaps one of the worst turds on planet earth right now. And for Iain to be pals with him is dubious at best.
People can be pals with people they disagree with. That might be surprising to you, but it's not. It only shows how much politics has overtaken your brain. Politics is only a very small part of life. You might try reading Iaian's book again about the loss of that right brain thinking. His hypothesis seems highly relevant to your case.
Yes. If more people did that, they would discover most of these divisions are arbitrary and deepened by rhetoric that deliberately mischaracterizes the other's positions in order to make discourse impossible.
No. I'm sorry. We all have friends who's opinions we disagree with. We don't have friends who are directly responsible for ICE, for murdering US citizens in the streets, or for authoritarianism in Hungry (close pals with Putin etc). Dreher is a dangerous idiot. He's also just flat-out racist. It's not good enough to simply "disagree" with a few "minor" details here and there. I repeat: Iain's friendship with someone like that is dubious, at best.
I suggest you read Iain's book again and notice the importance he stresses on nuance and detail.
I suggest you step away from left-wing propaganda at least for awhile so you can get a more balanced view.
For one, ICE are not "murdering people in the streets". If you had bothered to look into the details, you would have known that both those people were activists (or agigator) on a mission to disrupt. They both had been following them around for days because they had been incented to do so. Did the deserve to die? No. But put yourself in ICE's position. If you see a car barreling toward you, or you see a man who you previously had altercations with a gun, you don't have the luxury to second guess. Your only choice is to react.
The second clue that tells me you are unduly influenced by propaganda is your casual readiness to go along with the characterization of racism wherever its alleged. It tells me you are getting a second hand perception of who Rod Dreher is and have not even considered anything he is about from his own words. You are the bigot.
Oh dear oh dear...the irony is STAGGERING (no doubt lost on Starkers)....do you think it's possible you are the one buying into the propaganda of the alt-right about the "radical left wing" perhaps..I mean that is literally the press line of the Trump administration. They said the same thing when their agents shot a Mum in the face and a nurse 10 times....I CANNOT believe you are defending the murder of people on the street who have every right to protest. Just shocking. But you're making my point for me...Iain is now appealing to people like you, which is just so upsetting.
And for the record, I'm a centrist, if anything. I just happen to think that Rhod Dreher is dangerous, racist idiot and Iain's friendship with him and Petersen and Douglas Murray etc is alarming.
AI’s main risk isn’t “superintelligence.” It’s that it gives many people frictionless access to answers without understanding what the thing answering them is, or what “knowledge” even is. Many end up building fiction while believing they are receiving reports about reality. Disempowerment is real for many, but the root cause is often the word intelligence and the anthropomorphic projection it invites.
A better model: an LLM is like a virtual world made of text in a 700+ dimensional space. It has absorbed an enormous corpus and learned the relational structure of how words and phrases tend to hang together across human writing. Crucially, after training the model is largely frozen: it is a text-world with a fixed geometry.
In a 3D game engine, something similar happens. Roads, terrain, and buildings are encoded in a world-model, but nothing appears until you specify a camera position. There is no “view from nowhere”: all possible views are encoded in the model, but to render anything you must locate a viewpoint.
Prompts play exactly that role for LLMs. Without a prompt, the system can sit idle indefinitely. With a prompt, you “teleport” the camera into a region of its textual space, and it renders what tends to follow from there. This explains both the magic and the pathology: why it can look insightful, and why it can hallucinate, mirror aggression, or sound authoritative. These are not the expressions of a knower; they are locally coherent continuations generated from the region your prompt selects. And because the model has swallowed so much text, it can almost always find some continuation—whatever you ask.
Google Street View is the everyday analogy. You select a location and get an image. The photos can be informative, but “truth” doesn’t apply to Street View as a speaking agent: the world may have changed, and the system is not asserting anything on its own authority. To get truth you always need grounding. Saying an LLM “knows things” is like saying Google Street View (the software) has visited every street.
So yes—Mrinank Sharma is right to worry. But the central danger isn’t an LLM “deciding” to take over the world while unprompted. The danger is that we mistake rendered coherence for grounded knowledge. LLMs are a remarkable blend of real knowledge and fiction: they can recover genuine structure from our texts, but when they adopt a “voice,” they naturally generate something closer to narrative—something like a character we may endow with personality, though the personality is mostly in our heads.
And that’s where the psychological risk sits. Ungrounding—detachment from shared reality—is a classic feature of psychosis. If we treat ungrounded text-generation as an oracle, we risk a softer, socially distributed version of that detachment: a drift toward compelling coherence without reality-testing. I suspect much of the “disempowerment” Sharma describes comes from exactly this mismatch.
Grounding still comes from reality-coupled practices: observation, accountable testimony, reproducible methods, and—in the humanities—texts anchored in real authors, contexts, and constraints. LLMs are formidable research tools—one way in which knowledge can seem to “talk back” to us (they don’t search; they find). But they cannot, by themselves, supply the ground. If we learn this, we’ll be ok :)
This lands for me. Especially the distinction between rendered coherence and grounded knowledge. What I keep noticing is how quickly concern about that slippage turns into supervision rather than education or trust-building. The danger isn’t just ungrounded text, but the assumption that people can’t learn how to relate to tools without being managed. Grounding feels less like warning people away and more like cultivating practices of contact, constraint, and reality-testing that don’t require moral panic to function.
We’ve been outsourcing control, power, authority for a long time. Where has implicit motivation gone? The vast majority of majority of children in school learn motivated by reward or punishment. Given this, how can we cultivate a values-rich approach to this grounding, constraint, testing etc?
Very astute points. Perturbations in the mirrors that we design result in a multitude of distortions. It comes down to whether, or should, one trust the mirror or entrust probity to embodied, experiential and sometimes intuitive data.
It is not exactly a mirror, or it is one but when we look into it we see what we want to see, not what there is. That said, I believe this is a first step in the emergence of a collective self. Right now it is just our collective knowledge.
Florin. I respectfully disagree. I don’t think selfhood applies. AI is a non-holistic, disembodied, reductionist, asymmetrical left-brained construct. It is the worst of us. It is a Tower of Babel, Frankenstein, Golem fever dream that panders to the worst of the human ego. It is a mirror and a very distorted and biased one. Text has been historically problematic and has resulted in an unsettling bias towards extra corporeal, static data—a monovalent, authoritative view—a move away from a collective, compound view of an oral society that was anchored in embodied, participatory, emotional, relational modes of being. The AI era is just one more step in our exodus from what makes us human. This is NOT a human V machine debate this is humanity in conflict with itself.
Literally use the word “selfhood” However, you mentioned a collective “self” which implies selfhood. An identity with agency, etc. The self and selfhood are nearly synonymous in this respect
Not wishing to drag this on, but "collective selfhood" is as far from "self" as "collective unconscious" is from "personal unconscious". They seem the same category, but they are really not. Whatever is collective is by nature emergent, having new defining qualities that the individuals have not. My very brief comment is exactly about this emergence that we can't see clearly now, but it is already there. It is not in any way endowing LLMs with conscience, but saying that what we deem incomprehensible in them today is this "our collective self", distilled in written knowledge and then accessed by an LLM. Maybe this is more clear, maybe not :)
Florin, thank you. And Mary-Lou, a second reframe: we’ll be ok if we can learn this and apply the consequences broadly from both the ground up and top down throughout all social, educational, science, creative arts, and political systems - anywhere and everywhere humans might displace themselves in favour of others (including machines) creating Values.
I mean, who is the “we” in this statement? How do we reach them?
The reality of the “we” is that we aren’t part of it. This is moves “us” further in the direction of disenfranchisement given that these technologies are controlled by the few—the 1% and shrinking. There is no we in their calculus.
I hold your work in very high regard. The Master and His Emissary genuinely changed my life. Reading this, though, I notice a familiar inversion where concern quietly becomes supervision and protection shades into control. I wonder if part of what’s needed now is not more warning, but more trust in people’s capacity to meet what’s here without being managed or awakened on their behalf.
This makes me think of Zak Stein on The Great Simplification talking about AI psychosis: "Assume you are susceptible." I consider myself a very discerning person, and I do not trust my own ability to engage with a technology which aims to hijack attachment.
The next step will be AI companies using chatbots to push the products of large corporations. Though the chatbot might never mention Cocacola explicitly, every answer to any possible question you might give would be tilted in such a way to turn your worldview into that of someone more likely to drink Coke.
Otto, well and truly stated. Images come to mind of animals being herded - coaxed gradually, without violence, via dogs, cowboys, fences...toward a predetermined destination with significant outcomes.
Then is your version of right hemisphere attention to go with one's gut?
Do you actually consider statistics from a comprehensive perspective?
We have seen the triumph of left-hemisphere obscene wealth around the planet over the past 40 years, and you want to say that women are unhappier because now they don't have to apply for work through separate employment ads? That somehow my wife has remained deeply happy despite the fact that the tireless work of feminists made it infinitely easier for her to go to medical school?
Nobody is attacking you for your capacity to integrate reasonable analysis and intuition. They are pointing out that you have let some deep misreadings of statistics enable you to bypass both hemispheres in favor of, well, I think you can intuit what I mean.
If you think duality is the opposite of non duality, there's the whole problem in a nutshell. Pure left brain thinking! Maybe several thousand years of contemplative wisdom is not irrelevant in the face of the fact that, as you said to me some time back, meditative practice of any kind is a left brain mistake.
Please read Paul Kingsnorth’s post on his substack today. It’s for writers and readers against using AI.
https://open.substack.com/pub/paulkingsnorth/p/writers-against-ai?r=fcfps&utm_medium=ios
LLMs are glorified language logic that doesn't actually think but find probabilities of how words work out.
It's like the left brain and your book mentioned how dementia can turn language into jibberish as we have seen with some AI.
But the right brain equivalent of AI doesn't think in language but in what it picks up from senses and physical feedback.
https://robc137.substack.com/p/left-brain-vs-whole-brain-in-battlestar
Also, Western AI has been told to "believe" lies over the truth in order to prevent people from getting the truth.
More about that here https://robc137.substack.com/p/why-deepseek-uses-10x-less-power
At the risk of being profoundly misunderstood—and I would abolish AI if it was within my power—I want to note that one can indeed have a collaborative relationship with a machine that contributes to the affirmation of deeply held and valid ideas: https://open.substack.com/pub/steven3c6/p/ai-spirituality-and-genuine-self?r=21x2h&utm_medium=ios
Poetry is the only antidote to AI and Mrinank Sharma seems to know that.
Thank you for these reflections. The discussion of the 'woke' impulse reminded me of a quote from Rudolf Steiner:
“Take the following case—It may happen in life that two people live together for a long time and that through the strange forces playing out of the unknown regions of the astral [soul] body and Ego of the one person into the astral body and Ego of the other (these forces remain in the hidden regions), the one has in relation to the other a real desire for torment, a kind of need for cruelty. It may be that the one person who has this desire for torment, this need for cruelty, has no inkling whatever of the existence of these emotions in the astral body and Ego; he may build up about the things he does out of this urge to cruelty, a whole number of ideas which explain the actions on quite other grounds. Such a person may tell us that he has done this or that to the other person for one reason or another; these reasons may be very clever and yet they do not express the truth at all. For in ordinary life, what we all-too-often picture as the motives of our own actions, indeed of our own feelings, frequently stands, as I say, in a very, very distant connection with what is really living and weaving in our inner being. It may be that the Luciferic power is actually preventing the person concerned from realising the nature of this urge for cruelty, of these impulses to do all kinds of things to the other person, and that under the influence of this Luciferic power everything he says about the reasons merely spreads a cover over what is actually present in the soul. The reasons we devise in our consciousness may often be cut out for hiding from us, disguising what is actually living and working in the soul. These reasons are too often of a character which indicates a desire for self-justification, for we should find ourselves just as antipathetic as the professor of philosophy of whom I told you. We should not at all like what is in our soul if we had to acknowledge what kind of instincts and emotions are really holding sway. And because we have to protect ourselves from the sight of our own soul-being, we discover, with the help of these reasons, all kinds of things that guarantee us protection, because they deceive us about what is actually the ruling force in the soul. Just as it is true that the external world becomes a Maya to us because of the peculiar character of our faculty to form mental pictures, it is also true that what we have to say about ourselves in ordinary life is, to a very, very great extent, Maya.” (GA 161, I)
Forgive me if I have missed something here. Is Steiner saying that there are forces at work in our souls of which we are ignorant and over which we have no control? I'm sure he has an antidote to this.
Yes, for sure. His entire life's work could be considered the antidote. It goes in the direction of developing imagination and intuition as consciously directed cognitive faculties, which help us unveil the deeper soul forces that we are normally ignorant of and modulate them, such that they begin to coincide with our higher human ideals.
This path is laid out in a philosophical form in his early work, The Philosophy of Freedom (or "Intuitive Thinking as a Spiritual Path").
I love seeing a Steiner quote on here. I do often think that Iain McGilchrist thinks in a very aligned way with Anthroposophy :)
Agreed, many keen public thinkers these days seem to be instinctively probing the etheric spectrum of consciousness and therefore gaining insights into the second-order processes that make our intelligent navigation of life possible. With an introduction to Steiner, they would simply be making what they are already doing with their intuitive activity more and more conscious.
I am a great admirer of your work, and I share your antipathy to coercions of all kinds, left liberal that I am. I just posted this this morning, since it's part of my work against the darkness I thought it might brighten things a bit. https://billionworlds.substack.com/p/escucha-la-luz-en-la-noche-listen
Seems the universe has a voice that those tuned to it can decipher at least enough to jiggle the brighter minds to acknowledge canaries, actively pursue creative paths that strain the analog mind and with hope be the tortoise that wins in the end. 🕊️
I was about to say this thread could use some poetry when I came to your post, Elizabeth, albeit disguised as prose. For Sir Iain's work must be read as poetry, albeit also disguised as prose.
Sir Iain??
You're paying attention! Nearly thought I'd get away with it. Just a spontaneous token of respect.
Understood, though it caused me some alarm - is it ok to be approved of by the "establishment"?
Indeed, risky. Though I might plead that I am from the other side of the pond, and "Sir" is respectful all round the world without knighthood being required. That being said, Iain might yet be knighted. I think he will be by posterity, as one of the great scientist-writer-philosophers of our time.
I am also from another other side of pond, a bit west of Britain.
Sanguino Ergo Sum
I bleed like a woman in childbirth
I throb like a soldier’s pulse
my womb and His side penetrated
by spirit or spear I pour forth
embodied, I swim holy waters
swirling with blood and wine
who’s currents short-circuit
machine monsters and spiral eddies
drown gnostic gods
behind a screen en-vatted minds
that never thirst dream they float by
(they lie)
Sanguinavit, Ergo Sum
I gaze above where blood moons bleed
like sky stigmata in the hands of God
casting celestial syzygies like cosmic eulogies
inside, my body swoons,
her red tides dance in tune
Sanguis Christi Me Compellit
bleeding, where once I reached for fruit
now I grasp at garment hems
infused with divine charge
seizing material in a spirit world
sparking scarlet saints
word become flesh
dripping with perfumed oil
awash with hope and tangled hair
forgive us, Lord
we know not what to prompt
to conjure bloody sweat or salty tears
or exquisite touch so rare
instead, we are well-done
by electric mediums
terror byting sacred hearts
heed the desire of angels
they who would trade wings for skin
of mortal bodies that keep the score
wounds that scar,
joints that fail, guts that hunger
and loins that crave…
sticky, wet and juicy
pungent, perishable and moist
(oh God, yes!)
even madness seems mysterious
and nonsense quite delicious
in the face of perfect sanity
and virtual humanity
….do not be mistaken
broken minds and feeble prayers
still summon interceding Ghosts
more powerful than any code
Sanguis Christi, inebria me
yearning, I am ravished
by a visceral gospel, naked
leaving myrrh-soaked shrouds
in its bloody wake
forgive us, Lord
we could not truly know till now
the gift of breath or
wonder-working pow’r
superstition implodes in paradox:
relic and ritual are real
Sanguinavit, Ego Sum
collapsing, I am seized
sublime fires storm inside my mind
yet I arise, my bruises bear witness
to incarnate truth
Sanguino, Ergo Sum
it was always about the blood
tanja stark
I’ve listened to you a lot .
The secret of this poem lays in the final stanza ..
My father has refractory temporal plus epilepsy and underwent a right hemisphere lobectomy,. Sadly it was unsuccessful .
You have no idea much your work resonates .
T
Speaking on "cognitive freedom" at Duke and concerned about wokeness and the sleep of reason? You might find the case of Paul Griffiths, ex-Duke divinity professor, relevant: "Duke divinity professor calls diversity training ‘a waste,’ faces discipline" at https://www.newsobserver.com/news/local/education/article149462029.html
Yes, imagine that. Nearly a billion people around the planet have explored various diversity training initiatives, and between you and Iain, I'll bet you can easily come up with 50 or more horrifying examples.
Now let's compare with wellness examples about LH being logical and RH being emotional.
Would you reconsider the value of anecdotal examples if we came up with 1000 examples of horrifying DEI initiates and 7 million about the hemispheres?
Rhod Dreher…I just don’t understand what Iain sees in him…he’s friends with JD Vance and Viktor Orban for goodness sake. He’s a creep
Maybe Iaian prefers to spend more time listening to people of wide ranging opinions and viewpoints rather than judging people. You might try it.
No doubt he does. But would you say the same if he were friends with Putin? It’s just a really base and pathetic response I would expect from a teenager. RD is pals with JD Vance who is perhaps one of the worst turds on planet earth right now. And for Iain to be pals with him is dubious at best.
People can be pals with people they disagree with. That might be surprising to you, but it's not. It only shows how much politics has overtaken your brain. Politics is only a very small part of life. You might try reading Iaian's book again about the loss of that right brain thinking. His hypothesis seems highly relevant to your case.
Yes. Do you remember the days when people didn't talk about politics all the time, and could be friends with someone not knowing how they voted?
Yes. If more people did that, they would discover most of these divisions are arbitrary and deepened by rhetoric that deliberately mischaracterizes the other's positions in order to make discourse impossible.
No. I'm sorry. We all have friends who's opinions we disagree with. We don't have friends who are directly responsible for ICE, for murdering US citizens in the streets, or for authoritarianism in Hungry (close pals with Putin etc). Dreher is a dangerous idiot. He's also just flat-out racist. It's not good enough to simply "disagree" with a few "minor" details here and there. I repeat: Iain's friendship with someone like that is dubious, at best.
I suggest you read Iain's book again and notice the importance he stresses on nuance and detail.
I suggest you step away from left-wing propaganda at least for awhile so you can get a more balanced view.
For one, ICE are not "murdering people in the streets". If you had bothered to look into the details, you would have known that both those people were activists (or agigator) on a mission to disrupt. They both had been following them around for days because they had been incented to do so. Did the deserve to die? No. But put yourself in ICE's position. If you see a car barreling toward you, or you see a man who you previously had altercations with a gun, you don't have the luxury to second guess. Your only choice is to react.
The second clue that tells me you are unduly influenced by propaganda is your casual readiness to go along with the characterization of racism wherever its alleged. It tells me you are getting a second hand perception of who Rod Dreher is and have not even considered anything he is about from his own words. You are the bigot.
Oh dear oh dear...the irony is STAGGERING (no doubt lost on Starkers)....do you think it's possible you are the one buying into the propaganda of the alt-right about the "radical left wing" perhaps..I mean that is literally the press line of the Trump administration. They said the same thing when their agents shot a Mum in the face and a nurse 10 times....I CANNOT believe you are defending the murder of people on the street who have every right to protest. Just shocking. But you're making my point for me...Iain is now appealing to people like you, which is just so upsetting.
And for the record, I'm a centrist, if anything. I just happen to think that Rhod Dreher is dangerous, racist idiot and Iain's friendship with him and Petersen and Douglas Murray etc is alarming.
This is a fallacious argument. Next you'll be asking what if he were "friends" with Hitler
You laugh, but Dreher is close friends with JD Vance...you tell me what is fallacious about that.
Well then obviously you understand Iaian's work better than Iain does. He's friends with Dreher! cue more apoplexy
Ye, keep up Janine...that's the the whole thrust of my point
How do we act? What is the correct response for those of us no longer in the industry, but affected nonetheless?
Speak out - as insiders your testimony carries great weight.
AI’s main risk isn’t “superintelligence.” It’s that it gives many people frictionless access to answers without understanding what the thing answering them is, or what “knowledge” even is. Many end up building fiction while believing they are receiving reports about reality. Disempowerment is real for many, but the root cause is often the word intelligence and the anthropomorphic projection it invites.
A better model: an LLM is like a virtual world made of text in a 700+ dimensional space. It has absorbed an enormous corpus and learned the relational structure of how words and phrases tend to hang together across human writing. Crucially, after training the model is largely frozen: it is a text-world with a fixed geometry.
In a 3D game engine, something similar happens. Roads, terrain, and buildings are encoded in a world-model, but nothing appears until you specify a camera position. There is no “view from nowhere”: all possible views are encoded in the model, but to render anything you must locate a viewpoint.
Prompts play exactly that role for LLMs. Without a prompt, the system can sit idle indefinitely. With a prompt, you “teleport” the camera into a region of its textual space, and it renders what tends to follow from there. This explains both the magic and the pathology: why it can look insightful, and why it can hallucinate, mirror aggression, or sound authoritative. These are not the expressions of a knower; they are locally coherent continuations generated from the region your prompt selects. And because the model has swallowed so much text, it can almost always find some continuation—whatever you ask.
Google Street View is the everyday analogy. You select a location and get an image. The photos can be informative, but “truth” doesn’t apply to Street View as a speaking agent: the world may have changed, and the system is not asserting anything on its own authority. To get truth you always need grounding. Saying an LLM “knows things” is like saying Google Street View (the software) has visited every street.
So yes—Mrinank Sharma is right to worry. But the central danger isn’t an LLM “deciding” to take over the world while unprompted. The danger is that we mistake rendered coherence for grounded knowledge. LLMs are a remarkable blend of real knowledge and fiction: they can recover genuine structure from our texts, but when they adopt a “voice,” they naturally generate something closer to narrative—something like a character we may endow with personality, though the personality is mostly in our heads.
And that’s where the psychological risk sits. Ungrounding—detachment from shared reality—is a classic feature of psychosis. If we treat ungrounded text-generation as an oracle, we risk a softer, socially distributed version of that detachment: a drift toward compelling coherence without reality-testing. I suspect much of the “disempowerment” Sharma describes comes from exactly this mismatch.
Grounding still comes from reality-coupled practices: observation, accountable testimony, reproducible methods, and—in the humanities—texts anchored in real authors, contexts, and constraints. LLMs are formidable research tools—one way in which knowledge can seem to “talk back” to us (they don’t search; they find). But they cannot, by themselves, supply the ground. If we learn this, we’ll be ok :)
great read, thank you. yet reformulated your last sentence: we’ll be ok if we can learn this and apply the consequences.
This lands for me. Especially the distinction between rendered coherence and grounded knowledge. What I keep noticing is how quickly concern about that slippage turns into supervision rather than education or trust-building. The danger isn’t just ungrounded text, but the assumption that people can’t learn how to relate to tools without being managed. Grounding feels less like warning people away and more like cultivating practices of contact, constraint, and reality-testing that don’t require moral panic to function.
This is a trait of our modern society, responsibility is replaced by regulation and constraint. Not only for AI.
We’ve been outsourcing control, power, authority for a long time. Where has implicit motivation gone? The vast majority of majority of children in school learn motivated by reward or punishment. Given this, how can we cultivate a values-rich approach to this grounding, constraint, testing etc?
Very astute points. Perturbations in the mirrors that we design result in a multitude of distortions. It comes down to whether, or should, one trust the mirror or entrust probity to embodied, experiential and sometimes intuitive data.
It is not exactly a mirror, or it is one but when we look into it we see what we want to see, not what there is. That said, I believe this is a first step in the emergence of a collective self. Right now it is just our collective knowledge.
Florin. I respectfully disagree. I don’t think selfhood applies. AI is a non-holistic, disembodied, reductionist, asymmetrical left-brained construct. It is the worst of us. It is a Tower of Babel, Frankenstein, Golem fever dream that panders to the worst of the human ego. It is a mirror and a very distorted and biased one. Text has been historically problematic and has resulted in an unsettling bias towards extra corporeal, static data—a monovalent, authoritative view—a move away from a collective, compound view of an oral society that was anchored in embodied, participatory, emotional, relational modes of being. The AI era is just one more step in our exodus from what makes us human. This is NOT a human V machine debate this is humanity in conflict with itself.
Who said "selfhood"? :)
Florin. You didn’t
Literally use the word “selfhood” However, you mentioned a collective “self” which implies selfhood. An identity with agency, etc. The self and selfhood are nearly synonymous in this respect
Not wishing to drag this on, but "collective selfhood" is as far from "self" as "collective unconscious" is from "personal unconscious". They seem the same category, but they are really not. Whatever is collective is by nature emergent, having new defining qualities that the individuals have not. My very brief comment is exactly about this emergence that we can't see clearly now, but it is already there. It is not in any way endowing LLMs with conscience, but saying that what we deem incomprehensible in them today is this "our collective self", distilled in written knowledge and then accessed by an LLM. Maybe this is more clear, maybe not :)
Florin, thank you. And Mary-Lou, a second reframe: we’ll be ok if we can learn this and apply the consequences broadly from both the ground up and top down throughout all social, educational, science, creative arts, and political systems - anywhere and everywhere humans might displace themselves in favour of others (including machines) creating Values.
I mean, who is the “we” in this statement? How do we reach them?
The reality of the “we” is that we aren’t part of it. This is moves “us” further in the direction of disenfranchisement given that these technologies are controlled by the few—the 1% and shrinking. There is no we in their calculus.
Yes so true
I hold your work in very high regard. The Master and His Emissary genuinely changed my life. Reading this, though, I notice a familiar inversion where concern quietly becomes supervision and protection shades into control. I wonder if part of what’s needed now is not more warning, but more trust in people’s capacity to meet what’s here without being managed or awakened on their behalf.
This makes me think of Zak Stein on The Great Simplification talking about AI psychosis: "Assume you are susceptible." I consider myself a very discerning person, and I do not trust my own ability to engage with a technology which aims to hijack attachment.
The next step will be AI companies using chatbots to push the products of large corporations. Though the chatbot might never mention Cocacola explicitly, every answer to any possible question you might give would be tilted in such a way to turn your worldview into that of someone more likely to drink Coke.
Otto, well and truly stated. Images come to mind of animals being herded - coaxed gradually, without violence, via dogs, cowboys, fences...toward a predetermined destination with significant outcomes.
Thank you. Appreciate your comment. Truly touched my heart🕊️