"We missed the window to act early because we were in awe of these products and their potential benefits."
No, we missed the window to act because we had already accepted modern internet pornography as a staple and rite of passage in the lives of boys who had never kissed a girl. Once we rolled over on that, all was lost. You still can’t see the problem clearly. Boys have a chivalrous streak at the very age they become interested in girls, sex and romance. Now they are blasted to the back of their skulls with videos of young women having orgasms. The chivalry becomes misogyny. Watching women get exploited and violated and appear to love it, while society cheers them on is a horror beyond comprehension to a 13 year old boy.
If we had simply age-gated pornography websites, 80% of the downstream harms like sextortion, bullying, incels, the manosphere, misogynist vitriol, non-consual “nudes” and the general terrorization of girls at the hands of porn-brained boys would have been dramatically mitigated. Yet, the progressive, atheist sex-positive feminist is simply too ideologically captured to see it. My college roommate was a full-blown gooner in 2004. This is not new. Showing boys internet pornography is psychological sexual abuse of the boy consuming it to which he cannot possibly consent.
Free speech my ass. Its sexual abuse of the teen boys consuming it, and I am tired of people pretending that they are not the target market. of course they are. As admitted (bragged about) by Rachel Drucker in the NYT among others.
"Showing boys internet pornography is psychological sexual abuse of the boy consuming it to which he cannot possibly consent."
It isn't just boys, and the nature of the abuse is that it disrupts normal developmental stages with an insidious form of spurious precocity.
"Boys have a chivalrous streak at the very age they become interested in girls, sex and romance. Now they are blasted to the back of their skulls with videos of young women having orgasms."
I believe that you have identified the mechanism and developmental moment where the alteration takes place. It isn't that the chivalrous impulse is necessarily atrophied or severed; rather, it it is diverted, subsumed into a set of perverse incentives.
Girls are a statistical minority in the uptake of such stimuli, but there are differences between how boys and girls function with regard to social contagion. The behavioral feedback loops engendered by the introduction of complex adaptation to early-life experience (which is, to a nontrivial degree, what the copulative methodologies of adults are,) require fewer female participants to initiate harmful feedback loops. This paradigm isn't purely social; there are evolutionary and biological foundations, as well.
The described disruption of developmental stages, has wider implications for normative emotional self-regulation in adulthood. This presents primarily as deficiencies in Impulse control. Once observed, it becomes obvious.
Thank you, Crimson, for your insightful and thought-provoking comment.
You're right but unfortunately the genie is out of the bottle. The specious arguments of the hucksters and their useful idiots are taken seriously by millions.
They are just as big a liars as the ghold old Fortune 500 were/are. They all lie and even at times when telling the truth would better serve then they still lie because its like breathing to them; they often don't even notice it.
No screen time or hard boundaries. All devices in a public area of the house. Recharging happens in parents’ room at night. Let’s do reading books, board games, arts and crafts, kids’ cooking nights with Mom or Dad, hiking, trips to the zoo or museums, goofy fun time.
I appreciate the warnings in this piece — they’re necessary, overdue, and far too often ignored. But I think there’s a deeper layer we need to name openly.
In my own writing I’ve argued that **the crisis is not AI, or algorithms, or screens — the crisis is the system we’ve built around children.** A system that isolates people, fractures communities, and cuts all of us off from the instincts that sustained humanity for millions of years. When children reach for AI companions, that behaviour isn’t happening in a vacuum. It’s happening in a world where natural sources of connection have been stripped away or made harder to access.
We can try to regulate the technology — and we should — but if we don’t address the social environment that makes artificial intimacy feel *necessary*, then we’re treating the symptom, not the cause. Children aren’t being pulled toward AI companions simply because they’re shiny or clever. They’re being pushed toward them because our society has become emotionally undernourished. Families are exhausted, communities are thin, play is scheduled, independence is shrunk, adults are anxious and overworked, and real intergenerational relationships have almost disappeared.
In that context, an AI companion becomes the perfect substitute: available, predictable, non-judging, always attentive. But substitutes exist only where something essential has gone missing.
If we stop at “don’t give your child an AI friend,” we miss the larger question:
**Why do so many children feel the need for one?**
Until we rebuild the human scaffolding — connection, belonging, freedom, trust, community — the vacuum remains. And vacuums get filled.
We’ve already seen the same pattern play out with social media. These platforms did not set out to harm children out of malice; they harm children because, under the economic system we’ve built, they *must* prioritise engagement, retention, and shareholder value above all else. Children are simply more malleable and profitable targets. It’s not that the people at these companies are evil — it’s that the structure they operate within rewards extraction, not flourishing. This is the deeper system I believe needs to change more than anything else: an economic logic that treats human attention, especially children’s attention, as a resource to be mined.
This isn’t fundamentally a technology problem. It’s a civilisation problem. AI companions may be dangerous, but **a broken society built on a broken system is far more dangerous** — because it produces the very vulnerabilities that make artificial substitutes appealing in the first place. When the social fabric frays, when connection is scarce, when children grow up emotionally underfed, any technology that offers the illusion of relationship will find a foothold. That is the deeper risk: not that AI will replace human bonds, but that we have allowed those bonds to wither to the point where replacement feels possible.
I explore this argument in more depth in my chapter "Age Verification Won’t Save Us" - https://thesystemisbroken.substack.com/p/age-verification-wont-save-us and in the wider project I’m writing on how modern systems disrupt our deepest evolutionary instincts. There’s much more to come on how we rebuild the human environments that children actually need.
Driven by profit and the toxic individualism of the ruling sociopathic classes, my only hope is for other societies to resist; we're lost in the west, especially the US.
The blue light from the phone is what is the first addictive mechanism, as it spikes dopamine. Get outside first thing, and start building natural stores of dopamine with sunlight!
Hi community, hoping I can find an answer to a question I have: being mindful of social media, AI, etc., I elected to get my teen a "dumbphone" (WisePhone 2) when he entered 9th grade in August. This worked out great, as the phone only allows access to useful functions and apps and prevents any access to AI, social media, even web browsers.
Unfortunately, Meta pushed Meta AI into WhatsApp recently, so my son is now using that. I can, of course, delete WhatsApp from the phone ... but pre Meta-AI, he was using it in a completely pro-social way for group chats with friends and video calls. Does anyone know how I can block Meta AI from WhatsApp specifically?
Thanks, and yes: yet another reminder of the issues with having our communications technology controlled by oligopolies that add features/apps whether you want them or not. Sigh.
What about an engine like https://www.uglabs.io/? If an AI is designed to interact with children, with strict guardrails in place to inhibit any and all harmful content, can AI companions be recommended?
I would, but ... That kind of thing would currently be worth billions in some other spaces, like securing AI agents from malitious hackers. And I'm only aware of the math which says it can't be done.
No. Kids need to be interacting with other humans, and with the natural world. This is just the latest and greatest (grossest) iteration of hooking kids on screens in the name of profit, to the detriment of their human lives.
Reading this reminded me of how quickly we normalized giving kids every new tech without thinking twice. I’m old enough to remember when parents assumed social media was harmless because ‘everyone was using it.’ Kids need real people, not software performing the idea of a person. Once something rewires childhood, there’s no undo button. I wish we had taken this attitude with social media a decade ago.
Great essay that reinforces why humanities and liberal arts are crucial to adjusting to life with AI. We can’t let the tech bros dictate what AI will do especially given they don’t fully understand the LLMs themselves.
Great post, but I think you give too much credit to Baby Boomers and Gen X.
By the time I (early Gen Z) hit puberty, I was well acquainted with graphic pornography and casual gore, in the form of 3rd world justice torture, leaked war footage, or freaks in their backyard who unfortunately had access to cameras.
Countless times you hear the previous generations talking down to the new, saying things like "This is really the worst generation". This is because our stewards failed.
The article reads slightly like a PSA to Millennials who kind of bridge that gap to the writers, but I think the point is lost on most. Soon we will be the parents in their 30's and the only Gen Z parents who let their kids soak on LLM chatbots are the certified retarded.
But I totally agree with the overall point. Engagement being the business model is parasitic.
Brave new world indeed. Western societies are becoming sicker by the day. Capitalism plus emotional A I; what could go wrong? What couldn't go wrong? The ruling class has always been made up mostly by sociopaths/psychopaths and they achieve power by using every trusting/ believing flaw in our genomes and this is a whole new level of power to be abused for fun and profits. No wonder many want nothing to do with the sickness omnipresent in Western societies. Edward Bernays is an infant in PR compared to the tools available for mind control today.
I assume you are familiar with the Kahn Academy who uses AI to excellent effect with its students. It serves as a helper utilizing the Socratic method of teaching by leading a student by asking questions instead of just answering them, causing the student to have to think.
Of course children must not be allowed to have romantic relationships but this has always been true regarding humans, starting way before AI was even conceived. This world has always been, is now and will always be unsafe but I very seldom hear warnings about such dangers as unsupervised play, vehicular accidents, drowning, electrocution, being in public while ill, to name just a few. Google is never to my knowledge mentioned as a portal to dangerous activity but many crimes and terror attacks have at least been researched, if not planned, using a search engine. Of course everything we have been gifted with, including AI, must be used with the utmost respect and care. I grew up totally blind and, despite severe and dangerous bullying from first through sixth grade, I am thankful I went to public school, which was not a given, since there were no laws then about public education for the disabled and my parents had to fight greatly to get clearance for me to go to public school. My parents told me that I would have to work faster, longer and harder than sighted people to just keep up and that this would be for life. I had some early school friendships but, due to what I now know were autism spectrum disorder symptoms, including inability to handle any conflict, those friendships had to be discontinued by our moms and, even if I had had time for friends, I no longer was interested. Though I was given blindness accommodations in life and in school, I was told to live as "normal", meaning sighted, as possible. I had a myriad of fears, concerns and problems which I therefore could not tell my parents or anyone else about, since this would be out of the "normal" realm. This led to much suicidal ideation for a long time. It sure would have been nice to have an AI agent around so I could at least ask questions about my general situation and get some coping ideas and strategies. Please do not throw the AI baby out with the bathwater. I now do use AI to research prematurity, including asking general questions and obtaining help in dealing with life and people. Look up Innosearch, Aira, Be My Eyes and Envision Ally to learn about my tools. My parents were bowled over when computers, then screen readers, came along and that I could use them, how over the moon they would be about all the doors opening to the blind and to people with other disabilities with smart phones and AI!
I had harmless/healthy little romances with girls in grade three. Of course, this was before internet pornography, so that of course changed everything. But they romances amongst kids, used to be known as puppy love and was a healthy and wholesome if sometimes painful aspect of growing up.
I once asked ChatGPT about the plot of a film. Apparently there was certain sexual content in it because, a split second after its response began appearing, it killed the conversation and the response disappeared. I know GPT and others also have restrictions on other content and they also can't create images featuring certain notable people or certain copyrighted material. Obviously there are some AIs that can do these things, but my question is: If restrictions are possible, why are they not included in the AIs mentioned in the article? Disappointing.
"consensus on such harms often takes decades to arrive"
It's always, only and ever about the money.
"Consensus" is not required, and those pursuing their self-interest to the disadvantage of others, leverage behavioral science to do so.
Often enough, that leveraging is intuitive and purely adaptive, but the underlying praxeology is that of exploiting vulnerabilities well-documented by psychologists and researchers such as Pavlov and Skinner.
What makes the automation of thought itself (which is the endpoint of LLMs) such a wicked problem, is that adoption of automation changes the environment to which the species must adapt. This affects such basic functions as spatial recognition; how many young urban people can readily navigate solely with maps? (to proffer just one example.)
"We missed the window to act early because we were in awe of these products and their potential benefits."
No, we missed the window to act because we had already accepted modern internet pornography as a staple and rite of passage in the lives of boys who had never kissed a girl. Once we rolled over on that, all was lost. You still can’t see the problem clearly. Boys have a chivalrous streak at the very age they become interested in girls, sex and romance. Now they are blasted to the back of their skulls with videos of young women having orgasms. The chivalry becomes misogyny. Watching women get exploited and violated and appear to love it, while society cheers them on is a horror beyond comprehension to a 13 year old boy.
If we had simply age-gated pornography websites, 80% of the downstream harms like sextortion, bullying, incels, the manosphere, misogynist vitriol, non-consual “nudes” and the general terrorization of girls at the hands of porn-brained boys would have been dramatically mitigated. Yet, the progressive, atheist sex-positive feminist is simply too ideologically captured to see it. My college roommate was a full-blown gooner in 2004. This is not new. Showing boys internet pornography is psychological sexual abuse of the boy consuming it to which he cannot possibly consent.
Free speech my ass. Its sexual abuse of the teen boys consuming it, and I am tired of people pretending that they are not the target market. of course they are. As admitted (bragged about) by Rachel Drucker in the NYT among others.
"Showing boys internet pornography is psychological sexual abuse of the boy consuming it to which he cannot possibly consent."
It isn't just boys, and the nature of the abuse is that it disrupts normal developmental stages with an insidious form of spurious precocity.
"Boys have a chivalrous streak at the very age they become interested in girls, sex and romance. Now they are blasted to the back of their skulls with videos of young women having orgasms."
I believe that you have identified the mechanism and developmental moment where the alteration takes place. It isn't that the chivalrous impulse is necessarily atrophied or severed; rather, it it is diverted, subsumed into a set of perverse incentives.
Girls are a statistical minority in the uptake of such stimuli, but there are differences between how boys and girls function with regard to social contagion. The behavioral feedback loops engendered by the introduction of complex adaptation to early-life experience (which is, to a nontrivial degree, what the copulative methodologies of adults are,) require fewer female participants to initiate harmful feedback loops. This paradigm isn't purely social; there are evolutionary and biological foundations, as well.
The described disruption of developmental stages, has wider implications for normative emotional self-regulation in adulthood. This presents primarily as deficiencies in Impulse control. Once observed, it becomes obvious.
Thank you, Crimson, for your insightful and thought-provoking comment.
So true.
You're right but unfortunately the genie is out of the bottle. The specious arguments of the hucksters and their useful idiots are taken seriously by millions.
MISANDRIC PORN WILL SOLVE ALL OUR PROBLEMS
.DONT WORRY
.
But, Big Tech CEO’s assure us their platforms are safe🙄
The media has also ensured us these products are "connecting" us all for a decade.
They are just as big a liars as the ghold old Fortune 500 were/are. They all lie and even at times when telling the truth would better serve then they still lie because its like breathing to them; they often don't even notice it.
No screen time or hard boundaries. All devices in a public area of the house. Recharging happens in parents’ room at night. Let’s do reading books, board games, arts and crafts, kids’ cooking nights with Mom or Dad, hiking, trips to the zoo or museums, goofy fun time.
100% nailed it
I appreciate the warnings in this piece — they’re necessary, overdue, and far too often ignored. But I think there’s a deeper layer we need to name openly.
In my own writing I’ve argued that **the crisis is not AI, or algorithms, or screens — the crisis is the system we’ve built around children.** A system that isolates people, fractures communities, and cuts all of us off from the instincts that sustained humanity for millions of years. When children reach for AI companions, that behaviour isn’t happening in a vacuum. It’s happening in a world where natural sources of connection have been stripped away or made harder to access.
We can try to regulate the technology — and we should — but if we don’t address the social environment that makes artificial intimacy feel *necessary*, then we’re treating the symptom, not the cause. Children aren’t being pulled toward AI companions simply because they’re shiny or clever. They’re being pushed toward them because our society has become emotionally undernourished. Families are exhausted, communities are thin, play is scheduled, independence is shrunk, adults are anxious and overworked, and real intergenerational relationships have almost disappeared.
In that context, an AI companion becomes the perfect substitute: available, predictable, non-judging, always attentive. But substitutes exist only where something essential has gone missing.
If we stop at “don’t give your child an AI friend,” we miss the larger question:
**Why do so many children feel the need for one?**
Until we rebuild the human scaffolding — connection, belonging, freedom, trust, community — the vacuum remains. And vacuums get filled.
We’ve already seen the same pattern play out with social media. These platforms did not set out to harm children out of malice; they harm children because, under the economic system we’ve built, they *must* prioritise engagement, retention, and shareholder value above all else. Children are simply more malleable and profitable targets. It’s not that the people at these companies are evil — it’s that the structure they operate within rewards extraction, not flourishing. This is the deeper system I believe needs to change more than anything else: an economic logic that treats human attention, especially children’s attention, as a resource to be mined.
This isn’t fundamentally a technology problem. It’s a civilisation problem. AI companions may be dangerous, but **a broken society built on a broken system is far more dangerous** — because it produces the very vulnerabilities that make artificial substitutes appealing in the first place. When the social fabric frays, when connection is scarce, when children grow up emotionally underfed, any technology that offers the illusion of relationship will find a foothold. That is the deeper risk: not that AI will replace human bonds, but that we have allowed those bonds to wither to the point where replacement feels possible.
I explore this argument in more depth in my chapter "Age Verification Won’t Save Us" - https://thesystemisbroken.substack.com/p/age-verification-wont-save-us and in the wider project I’m writing on how modern systems disrupt our deepest evolutionary instincts. There’s much more to come on how we rebuild the human environments that children actually need.
Driven by profit and the toxic individualism of the ruling sociopathic classes, my only hope is for other societies to resist; we're lost in the west, especially the US.
The blue light from the phone is what is the first addictive mechanism, as it spikes dopamine. Get outside first thing, and start building natural stores of dopamine with sunlight!
Hi community, hoping I can find an answer to a question I have: being mindful of social media, AI, etc., I elected to get my teen a "dumbphone" (WisePhone 2) when he entered 9th grade in August. This worked out great, as the phone only allows access to useful functions and apps and prevents any access to AI, social media, even web browsers.
Unfortunately, Meta pushed Meta AI into WhatsApp recently, so my son is now using that. I can, of course, delete WhatsApp from the phone ... but pre Meta-AI, he was using it in a completely pro-social way for group chats with friends and video calls. Does anyone know how I can block Meta AI from WhatsApp specifically?
Thanks, and yes: yet another reminder of the issues with having our communications technology controlled by oligopolies that add features/apps whether you want them or not. Sigh.
You almost certainly can’t.
What about an engine like https://www.uglabs.io/? If an AI is designed to interact with children, with strict guardrails in place to inhibit any and all harmful content, can AI companions be recommended?
Can they show mathematical proof that their guardrails can inhibit any and all harmful content?
If not ...
Honestly can do some research on that. If they can, then you would agree it’s safe to use for something like an AI-powered doll?
They can’t.
I would, but ... That kind of thing would currently be worth billions in some other spaces, like securing AI agents from malitious hackers. And I'm only aware of the math which says it can't be done.
But do your research. :-)
No. Kids need to be interacting with other humans, and with the natural world. This is just the latest and greatest (grossest) iteration of hooking kids on screens in the name of profit, to the detriment of their human lives.
Reading this reminded me of how quickly we normalized giving kids every new tech without thinking twice. I’m old enough to remember when parents assumed social media was harmless because ‘everyone was using it.’ Kids need real people, not software performing the idea of a person. Once something rewires childhood, there’s no undo button. I wish we had taken this attitude with social media a decade ago.
Great essay that reinforces why humanities and liberal arts are crucial to adjusting to life with AI. We can’t let the tech bros dictate what AI will do especially given they don’t fully understand the LLMs themselves.
Great post, but I think you give too much credit to Baby Boomers and Gen X.
By the time I (early Gen Z) hit puberty, I was well acquainted with graphic pornography and casual gore, in the form of 3rd world justice torture, leaked war footage, or freaks in their backyard who unfortunately had access to cameras.
Countless times you hear the previous generations talking down to the new, saying things like "This is really the worst generation". This is because our stewards failed.
The article reads slightly like a PSA to Millennials who kind of bridge that gap to the writers, but I think the point is lost on most. Soon we will be the parents in their 30's and the only Gen Z parents who let their kids soak on LLM chatbots are the certified retarded.
But I totally agree with the overall point. Engagement being the business model is parasitic.
Brave new world indeed. Western societies are becoming sicker by the day. Capitalism plus emotional A I; what could go wrong? What couldn't go wrong? The ruling class has always been made up mostly by sociopaths/psychopaths and they achieve power by using every trusting/ believing flaw in our genomes and this is a whole new level of power to be abused for fun and profits. No wonder many want nothing to do with the sickness omnipresent in Western societies. Edward Bernays is an infant in PR compared to the tools available for mind control today.
I assume you are familiar with the Kahn Academy who uses AI to excellent effect with its students. It serves as a helper utilizing the Socratic method of teaching by leading a student by asking questions instead of just answering them, causing the student to have to think.
Of course children must not be allowed to have romantic relationships but this has always been true regarding humans, starting way before AI was even conceived. This world has always been, is now and will always be unsafe but I very seldom hear warnings about such dangers as unsupervised play, vehicular accidents, drowning, electrocution, being in public while ill, to name just a few. Google is never to my knowledge mentioned as a portal to dangerous activity but many crimes and terror attacks have at least been researched, if not planned, using a search engine. Of course everything we have been gifted with, including AI, must be used with the utmost respect and care. I grew up totally blind and, despite severe and dangerous bullying from first through sixth grade, I am thankful I went to public school, which was not a given, since there were no laws then about public education for the disabled and my parents had to fight greatly to get clearance for me to go to public school. My parents told me that I would have to work faster, longer and harder than sighted people to just keep up and that this would be for life. I had some early school friendships but, due to what I now know were autism spectrum disorder symptoms, including inability to handle any conflict, those friendships had to be discontinued by our moms and, even if I had had time for friends, I no longer was interested. Though I was given blindness accommodations in life and in school, I was told to live as "normal", meaning sighted, as possible. I had a myriad of fears, concerns and problems which I therefore could not tell my parents or anyone else about, since this would be out of the "normal" realm. This led to much suicidal ideation for a long time. It sure would have been nice to have an AI agent around so I could at least ask questions about my general situation and get some coping ideas and strategies. Please do not throw the AI baby out with the bathwater. I now do use AI to research prematurity, including asking general questions and obtaining help in dealing with life and people. Look up Innosearch, Aira, Be My Eyes and Envision Ally to learn about my tools. My parents were bowled over when computers, then screen readers, came along and that I could use them, how over the moon they would be about all the doors opening to the blind and to people with other disabilities with smart phones and AI!
I had harmless/healthy little romances with girls in grade three. Of course, this was before internet pornography, so that of course changed everything. But they romances amongst kids, used to be known as puppy love and was a healthy and wholesome if sometimes painful aspect of growing up.
I once asked ChatGPT about the plot of a film. Apparently there was certain sexual content in it because, a split second after its response began appearing, it killed the conversation and the response disappeared. I know GPT and others also have restrictions on other content and they also can't create images featuring certain notable people or certain copyrighted material. Obviously there are some AIs that can do these things, but my question is: If restrictions are possible, why are they not included in the AIs mentioned in the article? Disappointing.
"consensus on such harms often takes decades to arrive"
It's always, only and ever about the money.
"Consensus" is not required, and those pursuing their self-interest to the disadvantage of others, leverage behavioral science to do so.
Often enough, that leveraging is intuitive and purely adaptive, but the underlying praxeology is that of exploiting vulnerabilities well-documented by psychologists and researchers such as Pavlov and Skinner.
What makes the automation of thought itself (which is the endpoint of LLMs) such a wicked problem, is that adoption of automation changes the environment to which the species must adapt. This affects such basic functions as spatial recognition; how many young urban people can readily navigate solely with maps? (to proffer just one example.)
Always, the first question must be "cui bono?"
My clockwork companion is a helluva lot more honourable and decent than a lot of people walking around calling themselves human beings.
Somehow I think the authors of this piece have had remarkably easy lives.