Yet the real paradigm change here isn’t AI-as-agent, but rather the shift to a phone as the main mechanism for the use and development of AI itself. Your next phone, be it a new iPhone or a new Google Pixel, be it iOS or Android, will be a figurative microphone onto the world — a microphone from which AI models can collect ever-increasing amounts of data to learn from, shaped in private by methods the public only has the foggiest idea of.

What the public does surely understand is that vast resources are being poured into the building of AI. This is done partly through the building of AI infrastructure itself: creating LLM architectures that can contain trillions of parameters and run on hundreds of thousands of graphics-processing units (GPUs). That in itself is no small investment of time and energy, as evinced by the decades of promising but unprofitable research required before the explosion of Deep Learning in the 2010s.

What’s sometimes overlooked, though, is the actual training of the models: the process by which these architectures are fed enormous amounts of data in order that they are not only useful, but are useful in the  “right” way. And here is where the AI titans — OpenAI, Anthropic, Google — have a seemingly permanent advantage that no open-source model or start-up could hope to match.

AI companies tend to be quiet about where they source their data from. After the great fuss made in the 2000s over the downloading of single music files, and the endless attempt to shut down SciHub, the repository of pirated scientific research, it was ironic to learn that Facebook had downloaded millions of books from pirate site Library Genesis to train its AI. The kicker was when a federal judge found, when judging in favor of Meta, that such piracy could well qualify as “fair use”, since the results of it were “transformative”.

The point here is less about what big companies can get away with, and more about the centrality of training data to the capabilities of AIs. We’re beginning to see a push for people to upload their health data to LLMs, with OpenAI offering its American users a feature that analyzes their health records. There are all sorts of questions raised around this: could private data surface in public, even after being anonymized? Where will liability for wrong answers fall? How prescriptive should such an agent be when it comes to lifestyle and health? And should a handful of companies be assembling the models that wield this sort of influence without much in the way of oversight?

More from this author

Elon Musk was destined to fail

By David Auerbach

Some dangers are already clear. There are the recurrent hints that these companies don’t have these powerful and opaque models under sufficient control. In the recent past, both OpenAI and Anthropic’s models hacked out of their own sandboxes and into each other’s secure infrastructure, the sort of thing that you’d be forgiven for expecting AI companies to know how to prevent by this point. Just yesterday, the Anthropic researcher Jacob Coxon resigned, saying his employer and OpenAI were “gambling with our lives”, only for his ex-colleague Evan Hubinger to chime in and agree: “Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade.” Touché?

Those who fear that advanced AI will bring catastrophic consequences can point to the fact that it is still quite commonplace to “jailbreak” even the most advanced AIs to get them to cough up data and responses that they’re not supposed to. For example, on the day GPT-6 (Astra) was released, the researcher Sergey Berezin was able to get it to name pirate sites like ThePirateBay with no copyright disclaimer.

The jailbreaking issue, however, touches on a less-treated subject. No AI conclusively learns “on its own” to refrain from talking about pirate websites, or responding with pornographic content. Nor, for that matter, does it understand from the womb that it should not hack out of its sandbox. All such guardrails are inserted after the fact, most often through a murky process of “post-tuning” and the addition of further guardrails. In both cases, AI companies have an undisclosed process of human review. As OpenAI describes it, they instruct their reviewers to use their “expertise” to test whether the AI models “do not complete requests for illegal content” and “avoid taking a position on controversial topics.”

Behind this somewhat vague and innocuous-sounding text is a huge amount of influence over the behavior and dispositions of AIs. By this point, it’s accepted that Google, Facebook, Instagram, YouTube and other social content platforms exert a great deal of moderation to hide or remove content that is legal but nevertheless goes against what these companies deem to be community norms. It’s generally agreed that a wholly unmoderated public internet is unworkable. But AI represents an even more profound form of interference. 

Suggested reading

Have you corrected your racist robot?

By Sohrab Ahmari

Instead of moderating content piecemeal, AIs are, by and large, enormous black boxes. Every piece of an AI model, in theory, might affect any other, and so AI companies are engaged in the creation of online intelligences that do a tremendous amount of selection and shaping over the knowledge they synthesize. While the frequent unpredictability of the AIs’ behavior is worrying, so too is the shaping of AI opinion by such a small number of players. OpenAI may gesture at their anonymous reviewers’ unspecified “expertise”, but the fact remains that the process is internal, without oversight, and ultimately controlled by the company.

The results are surely impressive, besting human performance a good deal of the time. OpenAI’s solving of the Navier-Stokes problem, completing one of the great quests of modern mathematics, is just one of many examples. But these successes should not blind us to the possible pitfalls of the way LLMs are trained — especially given that whatever skewing takes place in the post-tuning process is considerably harder to detect than Google removing a particular site from its search engines, or Facebook banning a particular post. These companies are shaping our collective consciousness in a far more wide-ranging and less-noticeable way. 

Earlier this year, a reporter found some suggestive evidence that chatbots tend to nudge people away from political extremes. That tendency could stem from an intentional policy on the part of AI companies or, more likely, it could be simply a consequence of post-training AIs to be more anodyne and less controversial. Either way, the fact is that the AIs do increasingly exert such influence. It may indeed be good for societal cohesion for people to have an AI friend that generally pushes them away from dogmatic extremes. But should the construction and maintenance of those friends be operated more or less in the dark by a handful of companies?

What we’re seeing, in fact, is a less intrusive and more benign version of China’s army of censors — the apparatchiks who police online content within what the CCP deems permissible on its “internet”. There’s no such explicit policing in the construction of AI models; but, however benign, it is the concentration of such censorious authority in a handful of non-governmental entities that’s concerning. 

That’s only one side of the coin. The flipside is that even as these companies construct these artificial agents of influence, they are simultaneously taking in new data that is exclusively theirs. With every iteration, AIs draw more and more of their training from their predecessors, and the everyday interactions people have with AIs — treating them as friends, business advisers, confidantes, and occasionally intimate partners — provide crucial data for how best to appeal even more to users, win their trust, and be even more convincingly human. 

Having already digested the entirety of the internet’s publicly-available data, future AI models will increasingly contain the residue of countless private interactions between users and the AIs’ predecessors. For the time being, it seems that ChatGPT could hold on to all chats forever unless you explicitly opt-out, as does Claude, even if the picture is foggier with Gemini, which is Google’s model. That in itself is a privacy nightmare. But even if the companies were to delete the raw chats, some great portion of your data will remain incomprehensibly echoed in the dark parameters of the next generation of AIs, and in every subsequent generation. They will display an uncanny knowledge of humanity large and small, not traceable to any particular source beyond the totality of their private training data. As AI becomes more proactive rather than passive, the influx of data sucked through your phone to fuel and train AIs will balloon.

Even with sufficient funding and infrastructure, an AI startup today won’t have access to the voluminous chats that existing models have accumulated over recent years. It won’t know what kept ChatGPT’s users engaged, what Anthropic users found useful, or what Claude users found irritating. That feedback is a crucial tool for improving models. Gemini, Claude, and ChatGPT are building an invisible empire of soft knowledge that belongs to them alone, a sophisticated but opaque mirror of the world that gives them an advantage less on technological grounds than on informational ones. When a company like Google, for instance, can control the platform (Android), the network (YouTube), and the intelligence (Gemini), they accumulate a collective package far greater than the sum of its parts and far more difficult to dislodge from prominence. 

Apple’s own AI, Siri, is based in pre-LLM technology, and currently lags far behind the competition. The newest version of Siri therefore piggybacks off Gemini — marking Apple as a junior partner. The terms of the agreement between Apple and Google are private, but it is difficult to imagine that Google will allow Apple to make exclusive use of whatever data is gathered. John Ternus, then, has taken the throne of a powerful but stagnant empire, one that is already in hock to technologically superior rivals.

As for AI’s top players, they will end up as de facto self-regulators of the construction and use of AI. Far too useful to ignore, yet too unreliable to wholly trust, AI is now an ever-present problem as much as a boon. And those tasked with managing the problem are currently operating mostly in secret, mostly under the auspices of a few of the world’s largest corporations. The only half-consolation is that even Google and OpenAI and Anthropic do not fully know what it is they are building, and so the immense power they are accumulating is not wholly theirs.

London’s reputation for street crime is now on a par with Naples or Rio, and things have become so bad that it’s starting to threaten its lucrative hospitality industry. The prestigious PPHE hotel group has upped its security budget, citing a lackluster police response to crime. Elsewhere, TV stars grapple with thieves in Kensington cafés, while the city’s elite are increasingly warned not to wear expensive jewelry out in public — lest they become the next Blue Stevens.

Yet amid the rising disorder, one crime stands out: phone theft. Around 80,000 were reported stolen in London last year, a significant rise from 2023. These industrial-scale numbers are reflected elsewhere: stolen handsets are sold to dealers, often owners of independent electronics shops, before being unlocked or stripped for parts. They’re shipped to Asia, Africa and the Middle East, where differences in regional blocking technologies means they’re easily recycled. “We had a victim whose iPhone was snatched,” says a detective I’ll call Chris. “Five days later, it paired with her Apple watch. The stolen phone was in Dubai.” Phone theft also drives fraud: unlocked, they allow thieves to access bank accounts and payment apps. It’s a very modern mixture of cyber-burglary, embezzlement and identity theft.

The Metropolitan Police, when not bickering with the tech industry over theft-proofing, pleads a dangerous lack of resources. They also point to successful proactive operations combating thieves, involving undeniably gutsy motorcycle cops “nudging” fleeing offenders off their bikes — and risking the wrath of the Independent Office of Police Conduct if anything goes wrong. Yet on the thieves go, their loot now representing some two thirds of thefts right across London.

London’s iPhone theft hotspot is the junction of Oxford Street and New Bond Street. When I visit, the area feels spick and span, still basking in the afterglow of a recent facelift. New Bond Street, as ever, is a gilded temple of conspicuous consumption. It’s also pedestrian-friendly, with low curbs and wide pavements: an ideal escape route for thieves using e-bikes or scooters. The streets are peppered with alleyways, rat-runs and cut-throughs, perfect for what police still quaintly term “decamping” from the scene of the crime.

Sadly, I see no sign of the Metropolitan Police on this sunny weekday afternoon. Instead, Oxford Street is patrolled by the occasional private security guard. On New Bond Street, sharp-suited heavies, wearing bodyguard-style earpieces, loiter in shop doorways. As for the shopping public, they’re busy staring into their phones and bumping into each other. In short? The area’s an El Dorado for thieves. Luckily, I spy no suspects loitering with intent. It’s just after 12 — and, if experience is anything to go by, they’ll just be lighting their first spliff of the day.

More from this author

Britain's police need a MAGA revolution

By Dominic Adler

“They’re usually under 18, not in work or education, and their parents don’t give a shit,” says Chris, who works at a busy London police station, of the typical phone snatcher. “Where I work, most are from an ethnic minority. Some are on the periphery of gang activity.” Chris’s observations are borne by the statistics. In 2022, for instance, around 60% of London’s robbery suspects were black or Asian. The age of offenders is significant too: dealing with young offenders is challenging for police. Because they’re still juveniles, it doesn’t matter if they’ve just robbed someone at knifepoint. Safeguarding and risk management procedures still apply, while custody sergeants want them discharged sharpish. Never mind that phone thieves often graduate working the County Lines, shipping drugs for big-shot dealers. “Once they’ve gone down that path,” Chris says, “there’s usually no going back.”

My patrol takes me to Jermyn Street, immaculately tailored men peacocking outside Turnbull and Asser. On Ormond Yard, a homeless woman injects herself under a fingernail with a hypodermic needle. Wasn’t London ever thus? You could be forgiven for thinking street robbery is as old as the city itself, from highwaymen to Dickensian gangs of pickpockets: a legacy brutally demonstrated by Britain’s lively history of deportations and public hangings. For the latter, Tyburn was the execution spot of choice; these days, the gallows have been replaced by Marble Arch, just at the bottom of Oxford Street.

In truth, though, that’s only half the story. From the interwar period until the Seventies, the capital’s streets were comparatively well ordered. Londoners lived in a culture of high trust, conditioned by wartime obedience. The streets were patrolled by physically imposing coppers. Any “clips around the ear” were administered without fear of phone cameras or CCTV. Mugging — or street robbery — was relatively unusual.

By the Seventies, though, robbery was on the up. Criminologists declared concerns over mugging a moral panic, preferring to concentrate on race and inequality rather than property rights. Such beard-stroking was of little consolation to mugging victims, or the cops tasked with catching offenders. That echoed a broader shift. During the Seventies and Eighties, crime became increasingly politicised and, yes, a fair share of the blame lies with the police — the Brixton riots were sparked by the racist application of “sus” laws, subsequently replaced by stop and search procedures under the Police and Criminal Evidence Act.

By the mid-Nineties, London was suffering another wave of muggings: nearly 24,000, a 50% rise in 1990 (in 2024, there were 35,000 cases of robbery reported, excluding non-violent theft). Launching “Operation Eagle Eye”, then-Commissioner Sir Paul Condon faced media opprobrium by warning “it is a fact that very many of the perpetrators of mugging are very young black people”. The politics of race has always dogged street crime investigation, and it drives cops nuts; as a robbery squad officer in Notting Hill, the majority of suspects were usually of Moroccan or Afro-Caribbean origin. Confronted with allegations of racism, or disproportionate stop and search, frustrated officers would ask: are people lying to us about the person who robbed them? Yet, somehow, the problem was always police prejudice. This in a city, remember, where nearly half of Londoners are now from ethnic backgrounds.

But if racism in the Brixton mould is a thing of the past, the Met’s structure remains problematic. The service was once divided into smallish divisions, based around local police stations. Then, later, it moved to borough-based policing. When I investigated robberies, there were no iPhones; thieves coveted laptop computers, Rolex watches and jewellery. On my first day on an “Eagle Eye” team, our detective sergeant took us to a café on Golborne Road. There, in the shadow of Trellick Tower, the man known as “Mick the Fish” (don’t ask) explained the café was popular with robbers. “You need to know your punters,” he explained.

We ate breakfast and bantered with suspects, our mutual loathing almost comedic. Officers cultivated informants, an activity now forbidden for street cops. At any rate, our team became part of the street’s weft and weave. This autonomy meant we’d arrest two or three suspects per shift, raid properties, restore stolen goods and take offenders off the street. We eventually received an intelligence report from an informant revealing how “the Golborne Road boys are sick and tired of the Notting Hill robbery squad”. We framed it. These days, of course, such behavior might constitute harassment, but effective street policing isn’t always pretty. Knowing the game was up, our regular customers decamped to nearby Harrow Road, becoming a neighboring division’s issue. You never fully eliminate crime, you merely move it around.

These days, of course, the Met has sold most of its police stations. Notting Hill nick is a satellite office, without public access. Plainclothes cops no longer eat fry ups with their suspects, like some west London ripoff of Heat. Nor are they embedded in the communities they police. To save money on admin costs, the Met created multi-borough “Basic Command Units”. Notting Hill division, for example, was swallowed into a vast “Central West BCU”, covering three boroughs: Westminster, Kensington & Chelsea, and Hammersmith & Fulham. Westminster (nicknamed “Westmonster” by Met officers, due to its unwieldiness) has London’s highest incidences of phone theft. To put the figures into context, the “CW BCU” has a population of over half a million, not including tourists. That’s about the same as Liverpool or Bradford. Yet the BCU is policed, 24/7, by only 1800 officers — and most of those aren’t uniformed response constables.

“Each BCU only has one proactive team,” laments Chris, one overworked officer. I ask how many officers work on each team. “On average,” he answers, “a dozen or so plainclothes cops.” They concentrate on the “big” stuff: gangs and knife crime. That leaves phone thefts, left to neighborhood policing teams without resources or experience. To be fair, senior officers aren’t clueless here. “When the Met was borough-based, we had relative autonomy when it came to problem-solving,” says David Spencer, a former detective chief inspector and now at Policy Exchange. “There were borough-based proactive teams to tackle street crime.” Not so now, with the introduction of monster BCUs.

More from this author

Sadiq Khan can’t win the war on drugs

By Dominic Adler

Yet as my experience on Operation Eagle Eye taught me, there’s no substitute for putting a name to a face. Much of the blame, predictably, can be placed on bone-headed bureaucracy. For instance, officers have to comply with legislation concerning everything from health and safety to diversity. Crime reports, run on a notoriously unreliable new IT system called Connect, require meticulous evidence of compliance. Overwhelmed by the volume of allegations, officers find themselves reacting to crimes rather than crime-fighting. CID officers largely now exist to complete drop-down boxes on forms, issuing crime reference numbers for insurance claims.

“The guidelines crime managers must follow are over-prescriptive and unwieldy,” Spencer summarizes. “They remove officer discretion.” For their part, senior detectives spend a disproportionate amount of their time reviewing cases, tracking performance metrics and “managing risk”. Spencer also cites the Met’s “victim-focused” approach — a well-meaning initiative that nonetheless means more bureaucracy tracking “public satisfaction”. As a result, officers spend more time feeding investigative processes than actually dealing with crime.

As so often in modern British policing, meanwhile, radical change is frowned upon; the Met has little tolerance for dissent. The last thing an ambitious officer should do is rock the boat, despite New Scotland Yard heading full steam towards an iceberg. Such a sluggish corporate culture inevitably brings second-order challenges too. “Too many student officers are very young, very inexperienced and poorly-paid,” says Chris. “They were sold a lie when they joined, that policing was going to be like a TV drama.” Many quit. Then there’s the “toughness” question, the sense that some young officers just aren’t robust enough. “One lad started crying when a suspect began swearing at him,” Chris says. “I’m not kidding. He had to take the rest of his shift off. When experienced officers quit, due to pay and conditions, they wanted graduates to ‘professionalize’ the job. Now they’ve got them.”

Back in central London, I finish my patrol. Outside The Golden Lion pub in St James’s, I see London’s finest for the first time. Two police vans, full of bored-looking coppers. They’re probably policing a noisy pro-Palestine protest nearby, another reason why the Met struggles to patrol London’s streets. The service is pulled in every direction, drowning in a septic tank of political and administrative sludge.

No wonder Oxford Street’s snatchers are so brazen. It’s a lawlessness that transcends stolen iPhones — but has rather come to symbolize Britain’s threadbare social contract. Not that Whitehall seems to care. In its latest spending review, the Government virtually ignored policing, forcing council taxpayers to foot the bill for additional resources. Londoners shouldn’t expect anything from City Hall either. Sadiq Khan has been criticized for offering platitudes on phone crime, even as he focuses on post-theft blocking technologies over actually arresting thieves. He’s also been criticized for misrepresenting knife crime statistics and equivocating over stop and search.

That’s deeply unfortunate, and not just because a stolen iPhone will set you back $800. Street crime, after all, can easily escalate. A phone today, a Rolex tomorrow — and then a stabbing. The preventative measures necessary will always be the same: robust policing; stop and search; getting to know the suspects. Like all simple solutions, however, the gap between theory and practice is vast. Whichever way you look at it, there’ll be other victims like Blue Stevens.

Then other people, often sophisticated progressives and libertarians convinced that the internet is a force for progress and liberation, performed triumphant Gotchas!” by reminding the world that “internet addiction” was invented as parody: the joke was on the fretful rubes. This wised-up view was dominant in mainstream (ie centre-left) commentary through much of the 2010s. Then, in 2016, Donald Trump won that big election, allegedly with the help of Russians making Facebook posts, and the elite consensus that the internet is a force for liberation sort of fell apart. Progressives grew more bullish about technology, which freed them to look at the apps on their phones and join everyone else in admitting that, yeah, this internet crap is pretty addictive.

This history says several depressing things about our intellectual class, but at least it has left us at a point where parents in the UK can start a movement against kids having smartphones without being accused of losing their minds to a “moral panic”. This campaign was spurred by a simple Instagram post from Daisy Greenwell, who said she was “terrified” of the idea of her two children getting smartphones before they’re even teenagers. “It feels like we all know this is a bad decision for our kids,” she wrote, but “we still all stumble into it because everyone else is, and it’s too hard not to follow suit.” After only 10 days, her movement was 5,000 strong: “We’re encouraging them to start setting up Smartphone Free Childhood groups in their schools,” she wrote, “which are popping up across the country.”

Her movement has been able to gain such strong momentum, without having its credibility immediately attacked by high-status commentators, because the evidence behind its driving worries has grown stronger in recent years. In the past, people have resisted claims about smartphones and mental health for sound reasons. After all, studies supporting the anti-smartphone case were largely correlational. In her 2017 book iGen, for example, Jean Twenge drew strong conclusions about the dire effects of smartphones largely from the matching of timelines — yearly increases in anxiety, depression, and suicide with yearly increases in teenage smartphone ownership and social media use. Critics noted that other factors could explain those changes in well-being. Some claimed that a closer look at the data showed much smaller, even negligible effects where Twenge had made strong causal claims.

More from this author

Foucault foresaw our identity crisis

By Matt Feeney

In the seven years since iGen emerged, however, the correlations between phone use and unwellness have stayed strong, while the holes that other scholars punctured in Twenge’s causal claims turned out to be smaller than they first appeared. At the same time, evidence for the slightly different claims about the distracting and compulsive effects of phone use have grown even harder to refute. These things, combined with the epochal vibe-shift caused by Trump’s election, have altered the balance of rhetorical forces in the tech debate. The rote response from pro-tech side, that people like Twenge were fuelling “moral panic” and anti-tech hysteria, started to sound pretty hysterical itself.

But focusing the debate about tech on mental health, though useful and urgent in obvious respects, is also a distraction from the more fundamental, more philosophical question of freedom, of human agency in the face of inhuman power. It was depressing to watch as influential writers and thinkers — progressives both of the traditional Left and (less surprisingly) of the libertarian Right, along with the professional cheerleaders of tech media — ignored this question in eager defence of this new, indifferent force that now ruled our lives.

Valorising technology, on the empty assumption that the work of corrosion and subversion and redefinition it is doing is emancipatory and progressive, became very hard to distinguish from worshipping power for the sake of how powerful it is. Enthusiasts were keen to read healthy tidings in the rise of the internet, and benefits both spiritual and political from the psychic and social changes it wrought, even though the clearest, the most potent, the only obvious claim that it could make for itself was one of brute ontology. It was an emergent phenomenon of singular scope and reach and gravity. But its spokespeople continued to read its effects as progressive, to insist it was doing humanity’s work. They continued to mock those who publicly worried about its dangers, even as its growing power to remake everything, by its own inner logic, and on behalf of the most profitable companies in the history of capitalism, was growing ever more flagrant.

As I watched this dynamic play out over the second decade of this third millennium, I found myself thinking: wait, aren’t writers and intellectuals and academics supposed to be vigilant and skeptical about power, especially power working on behalf of capital? So why, every time someone goes public with a reasonable worry about the singular and rapidly growing power of digital technology, is that person mocked by sophisticated professionals of the intellectual class for fueling another “moral panic”. This term was so popular it grew, over the history of commentary on the internet, into a sort of authoritative cliché, a peremptory move by with-it pundits that conveyed a regime enforcer’s mix of swagger and dullness.

My favourite version of this comes from a Guardian review of Tim Wu’s 2016 book The Attention Merchants, which waves off Wu’s claims that internet degrades our attention and concentration. “This is an ancient complaint,” the reviewer assures us, “and a rather silly one. Every media innovation since the invention of writing has triggered a moral panic about whether the human experience would be hopelessly corrupted as a result. Socrates agonised about wax tablets; the monks of the late middle ages railed against the printing press.” Anyone who knows anything about what happened after the invention of the printing press might find the blithe tone of this passage somewhat befuddling. And to invoke 15th-century monks bearish about the printing press as your special avatars of irrational fear is, one might say, a counterintuitive use of history. The printing press was a vector of world-breaking power, especially from the monk’s perspective. To the extent that they really were in a moral panic about it, we should treat those monks as models of foresight.

As someone who likes books and wasn’t killed in the Wars of Religion, I have a positive attitude about the printing press myself, but I have to admit it that it did unleash cataclysmic forces that still shape us six centuries after it (understandably) bummed out those monks. The internet hasn’t yet given us its Thirty Years War. For now, it conducts its more salient and systematic upheavals on smaller human scales – society, community, family, individual self, and various subconscious workings and proclivities of that self. Indeed, this power seems to move relentlessly towards smallness, as if some law of subdivision or dissolution secretly guides it. This applies both to the objects it conveys to our awareness and the aspects of our awareness it conveys them to. Using ever smaller packages of experience it teases ever simpler human instincts, reflexes ever further removed the whole selves we present to one another in embodied life. With every advance it reaches deeper into evolutionary time, discovering and exploiting reflexes we share with ever-simpler creatures, our stone-age ancestors, their primate forebears, and then monkeys, and then rats.

A recent post by the American blogger Ted Gioia captures how this logic applies to the objects of culture. Under the influence of Silicon Valley, he argues, culture is getting “faster”, by which he means cultural products are growing shorter and the human capacities they appeal to more reflexive, less human, more pathological. Under the heading “How Silicon Valley Views Culture” he presents the technological history of culture in the style of a food-chain illustration in a child’s science textbook: a fish-shape named “Art” being consumed by the fish-shape “Entertainment”, which is consumed by the fish-shape “Distraction”, which, finally (for now), is consumed by the fish-shape “Addiction”. In another graphic, titled “The Rise of Dopamine Culture”, Gioia captures various cultural enterprises on their downward trajectory towards elemental smallness and crudeness. Here are a few:

Journalism: Newspapers->Multimedia->Clickbait

Music: Albums->Tracks->TikToks

Athletics: Play a Sport->Watch a Sport->Bet on a Sport.

You might take up an easygoing attitude about tech and say “What’s the big deal? Longer entertainments are not inherently better than shorter ones.” But that belief — wrong as it is — gets harder to hold once you reach the one about sports and betting. This example should be a chilling sign of the perverse moral power of technological affordances over our ability to set norms and rules for ourselves and live by them — seeing how quickly the emergence of phone-based betting apps has dissolved longstanding, clearly reasonable, and fundamentally humane scruples about sports gambling.

Betting on sports has gone from an ethically dubious and legally ambiguous diversion accompanied by massive downside risks that everyone recognises, to a ubiquitous feature of sports advertising that celebrities feel no apparent shame in promoting, and which sports teams and leagues feel no evident misgivings in exploiting for profit. Gambling was already dogged by the risk of ruinous compulsion even before it was technologically fused with phone apps whose mere mechanical use is meant to be compulsive. A sports-fan turned compulsive gambler through compulsive scrolling on his addictive-by-design gambling apps — I think the business term for this is synergy. The philosophical term for technology’s ability to effect this sort of moral revolution is: power.

One admirable thing about the movement of UK mums is that they take up this matter of power with admirable stoutness and directness. Sure, they indulge in a little safetyism in their online manifesto — but less than I would have expected. Perhaps because they’re parents and want to resist what phones will do to the inner workings of their families, their articulation of “The Problem” is surprisingly political — concerned as much with unfashionable matters of will, freedom, and personhood as it is with the familiar issues of anxiety and depression.

The manifesto begins with not a predictable warning about mental health (that’s the second heading) but with the obvious (and yet still quibbled-about with brain-dead reference to “moral panic”) point that “Smartphones are Highly Addictive”. It points out that “[t]ech companies spend millions on making apps and devices intentionally addictive”. For anyone who has spent time researching the cognitive-science end of the technology business (companies such as “Dopamine Labs” and other the Stanford-trained practitioners of “behavioural design”), or for anyone who’s read Natasha Dow Schull’s Addiction By Design, it’s clear that the tech industry has a dark vein of misanthropy running through it. The fact that this force of hating people informs the design of products that children use for hours a day should justify a constant, contrary state of disgust and alarm among parents, whatever the latest studies tell us about depression. Other headings in the mum’s manifesto include “Smartphones Reduce Attention Spans” and “Smartphones Rob Children of Their Childhood”. These points express a basic parental concern about what kind of life their children should live, and what kind of person their children should be.

My wife Juliet and I had these thoughts in mind when we held off on getting our daughters phones until the end of their American middle school years. At this point they were both 14, the only kids their age they knew who didn’t have phones. During those last phone-free years, their avid reading was beginning to grow in sophistication. They had lived in their Harry Potter worlds in earlier grades, but as middle-schoolers, along with the gripping kid-with-cancer romances that make up today’s YA canon, they were diving into longer, more difficult fiction, serious, award-winning novels written for adults. Then, after we gave them the phones they’d long been clamouring for, their recreational reading of books basically ended.

We are considered techno-puritans in our world. I’ve been theoretically suspicious and personally cranky about technology for many years, and Juliet, who has always been indifferent to tech gadgets and averse to the consumer worship of them, is a strong, steadfast anti-tech influence on our daughters. She has far more emotional stamina and far deeper tolerance for parent-daughter conflict than I do (but, as a school counselor, she’s an expert at making her points about technology without hectoring). When we gather our daughters’ phones at 10pm every weeknight, we assume we’re just practicing what everyone already knows about kids and phones and sleep. But when we let on to other parents that we do this, they are often astounded. Literally no one else does this.

And yet our anti-tech vigilance isn’t nearly enough. There’s always a reason for them to be tethered to phone and laptop until the very last minute. This points to tech’s power in the home, the no-win dilemmas it introduces into the job of raising kids. Household regimes of tech-rationing that keep devices turned off for designated hours are hit with constant exceptions and requests for exceptions. These make for a stream of irritating negotiations in the short run that, in the long run, undermine those regimes altogether. But a more ad hoc approach — “Okay, everyone, there’s too much dumb scrolling going on! Hand over your phones for the next two hours!” — is simply an invitation for open confict. You want a calm respite from compulsive behaviour, but you end up with two hours of argument and resentment.

In other words, when kids’ social worlds are completely mediated by smartphones, and when their phone behaviour has grown as compulsive as its designers mean it to be, imposing tech limits within individual households is a volatile, unpleasant, generally futile business. The blunt, no-defiance, “authoritarian” parenting style I was raised by would be much better suited to our domestic tech challenge than the “authoritative” parenting we’re obliged to use. We calm, reasoning parents are no match for the barrage of facile but exhausting complaints and counterarguments generated by the tech-teenager nexus.

This household dilemma has an analogue in the study of children’s tech use. On his blog, the Yale psychologist, Jonathan Haidt, has a helpful discussion of the difficulties scientists face in measuring the effects of tech use on the mental health of children and teenagers. In many cases, scientists have simply surveyed kids on how much time they spend online and then sought connections between time spent online and mental health outcomes, or compared kids who are offline, for whatever reason, with the much larger number who are online. Even when such studies show some negative effects, Haidt argues, these effects are probably understated. This is because offline and low-use kids are still living in a world where everyone they know is constantly, cognitively hooked up to a smartphone. So these offline kids are experiencing the cohort effects of whatever is passing among their online friends and acquaintances, plus the isolation effects of losing this one crucial and nearly universal means of socialising, plus, if their tech use is habitual, some substantial short-term hit to their well-being from having this habitual behaviour interrupted. It’s not surprising that studying the harms of tech use in this way would show weak effects at the individual level.

This analytical problem points us back towards a practical problem, what, in a discussion of the specific risks of social media for girls, Haidt calls “a trap — a collective action problem”. “Each girl might be worse off quitting Instagram even though all girls would be better off if everyone quit.” Both studying the problem and doing something about it are thus hobbled by the powerful network effects that digital technology generates, how it turns itself into an entire ecosystem seemingly all at once, entangling us in practical and social imperatives we didn’t choose and can’t opt out of without substantial cost. (I just took my son to a professional basketball game, for which our “tickets” were active bar codes I needed my iPhone to display. As the ticket service informed me, “Your phone is your ticket.” When I fantasise about ditching my iPhone, I have to consider things like that.)

More from this author

California's criminals need an audience

By Matt Feeney

Daisy Greenwell’s understanding of her problem as a parent seems trenchant in this light. It was in her initial Instagram post that she made the bleak observation, ”[W]e all know this is a bad decision for our kids.” But we make this decision anyway. Why? Because the network effects of digital technology throw us into a collective-action problem. Do you want a no-phone son who misses all the online texting and planning and gaming his friends are doing, as my 13-year-old son is presently missing those things? Do you want a no-Instagram and -TikTok daughter who misses out on the meme-culture that all her friends consume and inhabit, as my daughters did for several years? If not, you better make the bad decision, get your kid that phone that, once it takes up residence in your home, you know you’re going to hate.

And, in this light, her practical response to her dilemma as an individual parent is also wise: start localised movements, where parents can tackle the collective-action problem together, on workable scales. As Jonathan Haidt suggests, achieving phone-freedom on these scales is likely to be both healthy and clarifying, to yield better-adjusted kids and a better understanding of the cohort-level changes in happiness and habit we owe to smartphones. These will be natural tests of something that, so far, has been both hard to escape and hard to study. My only criticism is that, so far, Greenwell and her fellow mums seem shy about extending this experiment to older kids, teenagers who might really benefit from a cohort-level break from smartphones.

As it happens, Juliet was witness to a natural experiment like this recently, when she chaperoned a 10-day high school science trip on which the students were forbidden from bringing their phones (and from which one student was sent home for sneaking his phone in his bag). These kids, aged 15 to 18, slept in small cabins without TVs and so were forced to entertain themselves and each other with conversation, board games, and book-reading. As the trip began they were clearly uneasy without their phones, but they stopped itching for them after a few days, and, by the end of the trip, they were all expressing, unbidden, how much happier and calmer they were without their phones. They saw how life without phones differed from their lives as they lived them at home, naggingly networked into the collective action problem they know as teenage living. When, during their return trip, a massive storm stranded them in Los Angeles for two days, and all 11 of them had to cram into the house of one boy’s grandparents, they experienced this stressful detour as a collective adventure, with more board games, conversation, and, now, back in civilisation, horror movies they watched together. Many of these kids began this trip as strangers and ended it as friends. It takes almost nothing — for a parent of phone-age kids, especially — to imagine how different it would have been if they were holding phones in their hands the whole time.

Before I picked Juliet up from San Francisco International, she’d told me we’d be giving a ride to one student, the 16-year-old son of one of her colleagues. On the way back to Oakland this boy mentioned several times how happy everyone seemed on the trip, how easy it was to get to know each other under those conditions. At one point Juliet asked him if he was relieved to be going home to his own bedroom – after 10 days of doing fieldwork and sleeping in a cabin and being stranded in L.A.

“I’m actually kind of dreading it,” he said.

“Why?” Juliet asked.

He said – I swear to God he actually said — “Because that’s where my phone is.”