I have become increasingly convinced that high-speed internet represents one of the most consequential mistakes in modern human history. I do not mean that the internet has no value, that every technological development associated with it is harmful, or that life before broadband was idyllic. I mean something more specific. We built a communications system so fast, so convenient and eventually so omnipresent that it overwhelmed social arrangements which had developed around the physical limitations of human beings for thousands of years. Those limitations were usually treated as problems to be solved. Distance was inconvenient. Waiting was inefficient. Boredom was unpleasant. Forgetting was undesirable. Being unreachable was frustrating. Local scarcity limited choice. A question without an immediate answer seemed needlessly difficult. One by one, the network removed these constraints. We congratulated ourselves as each disappeared. We paid much less attention to the fact that human culture had formed around precisely these constraints, and that removing many of them simultaneously would alter the experience of being a person.
There is an important distinction between the early internet and what followed it. The internet of the 1990s and early 2000s was already transformative, but for most people it remained physically contained. There was a computer somewhere. You sat in front of it. You connected. You used the internet. Then you got up and did something else. The computer had a location and the activity had an ending. Even in 2005, when broadband was rapidly expanding, the International Telecommunication Union counted about 216 million fixed broadband subscriptions worldwide. Mobile broadband then expanded at an extraordinary rate: active subscriptions rose from about 268 million in 2007 to 5.3 billion in 2018. By 2025 the ITU estimated that six billion people, 74 per cent of humanity, were using the internet. (ITU) This was not simply the adoption of another appliance. It was the construction, within a small fraction of a human lifetime, of a global environment through which communication, commerce, work, entertainment, news, navigation, friendship, courtship, photography, music, administration and increasingly personal identity could all pass.
The crucial transition occurred when the internet escaped the computer. Broadband made continuous access practical. Wi-Fi removed the cable. Smartphones removed the desk. Mobile broadband removed the building. Social platforms supplied an endless reason to remain connected. Notifications supplied reasons to return. Algorithmic feeds removed the natural stopping point. Cloud services allowed work and personal information to follow us everywhere. The result was an inversion of the original relationship between humans and computers. Once, using a computer was a discrete activity within ordinary life. Increasingly, ordinary life became something conducted through a computer network. A person could technically put the device down at any moment, but the social and economic systems around that person no longer necessarily stopped with them. Employers, friends, relatives, schools, banks, businesses, governments and entertainment services remained available and potentially demanding attention at all hours. Research on permanent connectivity has found that the social expectation of availability can itself create conflicts around autonomy and personal goals. (OUP Academic)
This altered something very old. Human existence has always been organised through rituals, using the word in its broadest sense. Ritual includes formal religious ceremonies, but it also includes dinner, visiting, shopping, courtship, celebration, mourning, storytelling, travelling, listening to music, watching performances, maintaining friendships and marking the passage from one part of the day to another. Anthropological and psychological research has repeatedly connected collective ritual with bonding, cohesion and affiliation. Studies have found associations between participation in ritual activity and stronger social bonds, while experimental work has demonstrated ways in which synchronised collective activity can increase feelings of cohesion. (PubMed Central (PMC)) These rituals were never separate from technological conditions. They developed within a world in which bodies occupied one place at a time, communication took time, information was incomplete and human attention was limited by what could physically reach it.
Consider how much ordinary life once depended upon transitions. Going to a cinema required leaving home, travelling somewhere, arriving at a specified time, sitting among other people, watching something that could not be paused and then physically leaving when it finished. Buying an album involved obtaining an object, carrying it home, opening it and deciding to play it. Calling someone meant approaching a telephone. If nobody answered, the interaction did not happen. Meeting someone required agreeing upon a place and trusting that both people would arrive. Travelling somewhere unfamiliar involved preparation, observation, asking people for directions and occasionally becoming lost. Watching television meant choosing among whatever was being broadcast at the time. A newspaper had an end. A magazine had a final page. A record finished. A shop closed. The evening eventually ran out of available things to do. None of this guaranteed wisdom or happiness. It created boundaries. Boundaries gave activities shape, and the existence of an ending allowed attention to return to the immediate physical world.
High-speed internet has progressively dissolved those endings. The modern feed has no final page because a final page would be economically irrational. Streaming libraries have no meaningful closing time. News arrives continuously. Messages accumulate while we sleep. Someone can always be contacted. Something can always be searched. There is always another video, another argument, another recommendation, another notification, another product, another photograph, another person speaking. The problem is not simply volume. Human beings have always lived among more information than any individual could absorb. The distinctive feature of the network is the removal of the physical barriers that once protected us from most of it. Researchers examining collective attention across several cultural domains found evidence that popular subjects rise and decay increasingly rapidly as the production and consumption of information accelerate. Their model described an environment in which a growing quantity of material competes for the same finite supply of human attention. (Nature) We did not expand human attention to accommodate the internet. We constructed an information system capable of making effectively unlimited claims upon an attention system that remains stubbornly finite.
Boredom is one of the casualties I think we have seriously underestimated. For most of human history, there were portions of life during which nothing particularly stimulating was available. A person waited for a bus. Sat in a waiting room. Looked through a train window. Lay awake. Stood in a queue. Arrived early. Walked somewhere. Waited for another person. These intervals were not necessarily enjoyable, but they were psychologically distinct territory. Thoughts wandered. The environment intruded. People watched other people. Memories appeared without invitation. Ideas developed because nothing more interesting immediately displaced them. Today those gaps can be filled within seconds. The hand goes to the phone almost before conscious boredom registers. The capacity to abolish an idle minute sounds trivial until it is repeated across billions of people, thousands of times, throughout years of life. A culture can lose an enormous quantity of unstructured mental time without any individual moment appearing important enough to mourn.
Privacy has undergone a related change. Previous generations certainly lived under intense forms of social scrutiny, particularly within small communities, families and institutions. What they did not routinely experience was the possibility that ordinary life could be documented, distributed, measured, retrieved and evaluated by enormous invisible audiences. Childhood photographs existed, but childhood itself was rarely an archive. An embarrassing conversation could disappear. A failed relationship could recede. A stupid remark could die with the afternoon in which it was spoken. An adolescent could experiment with an identity without necessarily producing a permanent searchable record of the experiment. The modern network encourages the conversion of experience into information. Photographs become uploads. Friendship acquires visible signals. Popularity acquires numbers. Conversation produces timestamps, activity indicators and read receipts. Personal taste becomes behavioural data. Location becomes data. Attention becomes data. Human tendencies that are ancient—status seeking, comparison, belonging, envy, curiosity, attraction, approval—now operate inside systems capable of recording and quantifying them continuously.
The scale of exposure has become especially clear among people who have never known another arrangement. Pew Research Center found in 2024 that 46 per cent of American teenagers described themselves as online almost constantly, roughly twice the proportion recorded a decade earlier, while 96 per cent said they used the internet every day. (Pew Research Center) The precise psychological consequences remain contested and should not be reduced to a single statistic. The OECD’s reviews of digital life emphasise a complicated body of evidence containing genuine benefits as well as risks, with outcomes varying according to the activity, person and context. (OECD) Even so, “almost constantly” is an extraordinary phrase when considered historically. It describes a condition in which connection to a global information network is no longer an event. It is the default state from which temporary disconnection becomes the exception.
This is also why discussions framed narrowly around “screen time” often feel inadequate. A screen is only the visible surface. The deeper transformation concerns the reorganisation of behaviour around perpetual connectivity. Two hours spent writing, drawing, talking to a distant relative or watching a film cannot sensibly be treated as psychologically identical to two hours of rapidly switching among notifications, short videos, advertisements and social comparison. Research itself increasingly reflects this distinction. A 2025 experience-sampling study involving more than 10,000 reported social interactions found substantial differences according to the form and context of digitally mediated communication: phone calls were rated highly, video calls could perform extremely well under some circumstances, while text and social-media interactions produced different outcomes depending upon relationship and purpose. (OUP Academic) Technology is capable of sustaining genuine relationships. It can connect families across continents, provide communities to isolated people, widen access to knowledge, enable telehealth, allow disabled people forms of participation previously denied to them and make communication extraordinarily inexpensive. Those achievements are real. They do not erase the structural problem created when virtually every useful digital service is delivered through the same infrastructure as an industrial system built to capture attention.
The danger lies partly in convergence. Earlier technologies usually occupied identifiable territories. The telephone handled calls. The television displayed programmes. The stereo played music. The camera took photographs. The map helped with navigation. The newspaper carried news. The typewriter produced documents. The game console played games. The shop sold goods. None of these categories was absolute, but their separation imposed useful friction. A person could engage with one activity without automatically entering all the others. The smartphone collapsed these territories onto a single object and the network connected that object to systems with effectively unlimited depth. The same movement of the thumb can take a person from a private conversation to a war, an advertisement, a stranger’s holiday, a workplace dispute, pornography, political propaganda, a family photograph, a financial account and a comedy video within minutes. Human psychology was not designed by anyone with this sequence in mind. It is simply what became possible, profitable and normal.
Culture has changed accordingly. Scarcity once forced cultural experiences to develop slowly. You might hear an album repeatedly because you owned six albums rather than having access to sixty million. A film could sit in your mind because the next film was not immediately available. Television created shared cultural schedules because large numbers of people watched the same thing at approximately the same time. Local shops, radio stations, newspapers, clubs and venues exerted disproportionate influence because geography constrained participation. High-speed internet vastly increased access and diversity, an enormous benefit in many respects, while simultaneously creating a culture in which almost everything competes with almost everything else. A difficult novel competes with a fifteen-second clip. A conversation with your partner competes with every notification arriving from outside the room. A local musician competes with the complete recorded history of popular music. An evening with your own thoughts competes with a machine that can generate stimulation indefinitely. The choice sounds liberating when described item by item. In aggregate it creates an unprecedented contest for consciousness.
I think this is why footage from roughly 1998 to 2008 can produce such an unusual sensation now. The period is recent enough to contain recognisably modern technology and distant enough to reveal how radically our relationship with it has changed. Watch an old technology programme, a game show, a convention recording or a piece of ordinary home video. Computers are everywhere. The internet is discussed constantly. Mobile phones exist. Online gaming is becoming normal. Digital culture is already forming. Yet the machinery still appears to occupy a place within life. A computer sits on a desk. A television belongs to a room. A camera appears because somebody has decided to film something. People can leave these objects behind. The technology has edges. By the end of the following decade, those edges have largely disappeared. The ITU’s figures capture the speed of the structural change: active mobile-broadband subscriptions went from hundreds of millions in 2007 to billions little more than a decade later. (ITU) A historical transformation of everyday human behaviour occurred with astonishing speed.
Calling this nostalgia misses the point. Every period contains suffering, stupidity and injustice, and there is no serious argument for reconstructing the social order of 1998. The argument concerns technological conditions. A society can become materially richer, medically better, more informed and technologically capable while simultaneously damaging important features of ordinary experience. Progress does not move along a single axis. High-speed internet solved genuine problems and created genuine opportunities. It also removed constraints whose social function we barely understood because nobody had previously possessed the ability to remove them. We discovered what happens when communication becomes instantaneous by making communication instantaneous. We discovered what happens when entertainment becomes inexhaustible by making it inexhaustible. We discovered what happens when children can remain connected to their peers throughout the entire day by giving them the means to do so. We discovered what happens when commercial systems can measure human attention at enormous scale by building those systems first and examining their consequences afterwards.
I use the word “doomed” deliberately, although I do not mean that human extinction is inevitable or that individual people are powerless. I mean that something has been lost which cannot simply be restored through better personal habits. The old arrangement depended upon everyone sharing limitations. You could be unreachable because being unreachable was normal. A shop could close because commerce closed for the evening. A person could answer a letter after several days because several days was an ordinary interval. Children could disappear into neighbourhood life because continuous remote supervision was technically impossible. Friends could lose contact because maintaining every relationship indefinitely required effort. Cultural experiences could remain scarce because distribution was physically constrained. Once the surrounding society reorganises itself around permanent connection, an individual who rejects that connection does not return to the previous world. He enters the current world with one of its principal infrastructures removed.
That is why “digital detox” has always seemed to me too small a concept for the scale of the problem. An individual can switch off a phone. An individual cannot restore a society in which nobody expects immediate replies. A family can put devices away during dinner. It cannot recreate a childhood culture in which every other child is also offline. A person can buy records, use paper maps, read printed books and refuse social platforms. Those choices may substantially improve a life, but they occur inside institutions, workplaces and communities already structured around networks. The disappearance was collective, which means the reconstruction would also have to be collective. We would need to decide that some forms of friction are desirable, that some inefficiencies protect human life, that some periods should remain unavailable to commerce and communication, that children deserve large territories beyond algorithmic systems, that not every moment should be documented, that not every question needs immediate resolution, and that the capacity to reach someone does not create an entitlement to their attention.
There is no realistic prospect of abolishing high-speed internet, nor would I advocate doing so. Too much essential infrastructure now depends upon it, and too many of its benefits are genuine. The useful question is whether we can become mature enough as a civilisation to recognise that maximum connectivity is not synonymous with maximum human flourishing. The first decades of broadband were governed by an assumption that faster, easier, more connected and more available were naturally improvements. We now have enough experience to question that assumption. The OECD has reached a similarly broader framing in its work on digital well-being, treating digital transformation as something whose effects extend across multiple dimensions of life rather than as a narrow matter of access or economic productivity. (OECD) Access solved one problem. What we do with permanent access is another.
For thousands of years, human beings lived within a basic fact that no technology had seriously overturned: most experience occurred among embodied people occupying finite places for finite periods of time. Information moved more slowly than thought. Distance created separation. Night interrupted activity. Objects remained where they were left. Other people eventually went away. Most moments disappeared after they happened. Culture adapted endlessly while those underlying conditions remained surprisingly durable. High-speed internet broke several of them within one generation.
We are still human. Our bodies remain local, our attention remains finite, our relationships still require time, and our children still arrive with nervous systems shaped by an evolutionary history unimaginably older than broadband. The network does not share those limitations. It is global, continuous, searchable, reproducible, measurable and effectively inexhaustible. We have spent twenty years attempting to make ourselves compatible with it.
I increasingly think we should have been asking whether it was compatible with us.
