Attention

The Cocktail-Party Problem

In 1953, the British psychologist Colin Cherry described a phenomenon that became one of the foundational observations of modern attention research. At a noisy cocktail party with many simultaneous conversations, you can focus on the conversation in front of you while screening out everything else. You hear the words your conversation partner says. You miss most of what is said two metres away. But if someone across the room says your name, you notice immediately. Your attention was not actually filtering everything else out completely; some lower level of your brain was monitoring all the conversations and would interrupt your conscious focus when something specifically relevant arrived.

What this revealed about attention. There is a top-down system that consciously focuses on what you have decided is important, and a bottom-up system that monitors the broader environment for things that might be important even when you are not consciously attending. The two systems operate in parallel. The bottom-up system can override the top-down system - your attention gets pulled by sudden movement, by your own name, by sounds that signal danger or opportunity. Modern attention research has substantially extended Cherry's observation but kept the basic two-system framework.

Why this matters in the modern environment. The bottom-up attention-grabbing system was designed for an environment in which sudden movement, a name being called, or a loud noise was likely to be relevant - signals worth interrupting current focus to investigate. The modern environment is full of artificial signals engineered to trigger the same system: notification sounds, badge counts, auto-playing video, push alerts, attention-grabbing thumbnails. Each one exploits the same neural mechanisms Cherry described, but in service of business interests that are not aligned with the attender's own goals. The attention-capture industry has become better at producing these signals over the last fifteen years; the underlying brain has not changed.

Top-down focus and bottom-up monitoring run in parallel - and the bottom-up wins more often than is comfortable

The Attention Economy

The idea was stated first and most compactly by Herbert Simon in 1971: a wealth of information creates a poverty of attention. In 1997 Michael Goldhaber applied it to the emerging internet, arguing that the digital economy would be characterised not by scarcity of information (information was becoming abundant) but by scarcity of attention. The total amount of attention available - the number of waking, conscious, attentive hours that humans collectively can give to anything - is roughly fixed at a population level. The amount of content competing for that attention has grown exponentially. The result was the development of what is now called the "attention economy" - businesses that compete to capture and monetise attention, advertising-supported platforms, content optimised for engagement rather than information value.

What attention is worth. Major platforms calculate revenue per user-hour as a key metric. Meta's revenue per user-hour, YouTube's, TikTok's, X's are all functions of how much attention the platform captures multiplied by how efficiently it can monetise that attention through advertising. In aggregate, the attention economy is roughly the size of the global advertising market, which passed $1 trillion for the first time in 2024 and is forecast at somewhere between $1.2 and $1.3 trillion for 2026 depending on whose forecast you take - though its true scale is larger because of the time-on-platform that supports the advertising ecosystem.

The structural consequences. The competition for attention is the underlying force shaping much of modern media, politics, and consumer technology. Platforms compete on attention; the products that win attention compete more aggressively for it; the design patterns that capture attention spread; the businesses that don't capture attention struggle. Over twenty years of this competition, the average product has become substantially better at capturing attention - which from the user's perspective often means substantially worse at respecting it. The modern smartphone interface, the average social-media feed, and the average online article all carry the design legacy of this competition.

The asymmetry that matters. The attention economy is built around capturing attention with substantially more sophistication than users have for protecting it. The platforms employ teams of designers, engineers, and behavioural scientists optimising for engagement; the typical user has whatever willpower they brought to the interaction. This asymmetry is not a moral judgment of either side; it is a description of how the system is designed. Recognising it is the first step toward thinking clearly about how to navigate it.

Attention is the actual product the modern advertising economy is built on

Inattentional Blindness

One of the most striking findings in modern attention research is "inattentional blindness" - when you are concentrating on one thing, you can fail to notice obvious things happening directly in your visual field. The famous Simons and Chabris "invisible gorilla" study had participants count basketball passes in a video; about half failed to notice a person in a gorilla costume walking through the middle of the scene. The study has been replicated many times; the effect is robust.

What this implies. Conscious perception is not "see everything in the visual field." It is "see what attention has been directed at; ignore the rest." When attention is fully engaged on a specific task, the brain does not waste resources processing things that are not relevant to that task, even when those things are large and obvious. The strength of the effect surprises people; people consistently overestimate how much of their visual field they actually consciously perceive at any moment.

The driving implication. Phone use while driving is so dangerous partly because the conversation engages the brain's attention systems even when the eyes are pointing forward. Drivers using phones report seeing cars and pedestrians they later turn out not to have processed at all - the visual information arrived but attention was elsewhere, so it never reached conscious processing. The legal restrictions on phone use while driving reflect this research; the day-to-day ignoring of those restrictions reflects how poorly most people understand the underlying mechanism.

The broader implication. Most of life happens through this filter. Things you are not attending to mostly do not register at all. Conversations you tune out, scenery you walk past, the body language of people you are with while you are checking your phone - all of these go through the filter without ever being consciously perceived. The "I was there but I do not remember it" experience is exactly what attention researchers would predict for moments of divided attention. The question of what you actually want to be present for, in your own life, is partly a question about where your attention has been.

When attention is fully engaged, the rest of the visual field stops registering

The Cost of Switching

The standard finding from cognitive-psychology research on task-switching is that switching between tasks has substantial costs in time, accuracy, and depth of engagement. Switching from writing to checking email and back to writing typically costs more than the few seconds the switch itself appears to take. The brain has to load the context of the new task, process whatever it finds, and then reload the original context when switching back. The reload is rarely complete. Each switch leaves fragments of the previous task active in working memory, which interfere with the new task.

What the research consistently shows. Studies measuring attention residue, task-switch costs, and productivity find that "multitasking" - rapidly alternating among several tasks - is substantially worse than focused attention on one task at a time, on most measures of output quality and total work completed. The exceptions are tasks that genuinely run in parallel (talking on a hands-free phone while doing routine driving) or tasks that are routine enough not to require focused attention. For anything requiring genuine cognitive engagement, single-tasking is measurably more effective than multitasking.

How this plays out in modern work. Gloria Mark's group at UC Irvine has been measuring this with computer logging rather than self-report since 2004, which makes it one of the few places where the trend itself is documented rather than asserted. In 2004 workers averaged about two and a half minutes on a screen before switching. By the late 2010s the same measurement gave about 47 seconds, a figure since reproduced by other groups in workplaces, among students, and at home. Her earlier work also produced the widely-quoted finding that returning to an interrupted task takes over twenty minutes on average - worth handling carefully, because it measures the gap until the original task is resumed, during which other work is being done, not twenty minutes of pure loss. The economic cost is real; the experiential cost of a working life spent in 47-second fragments is harder to measure and harder to dismiss.

What helps. Batching similar tasks (handling all email at scheduled times rather than continuously). Time-blocking (committing to focus on one task for a defined period without switching). Removing notifications during focus blocks. Working in environments where the structural pressure to switch is reduced. Each implements the cognitive-science finding that focused attention produces better work than fragmented attention. The reason these techniques are not standard practice is mostly that the modern environment makes them effortful to implement, not that they do not work.

Each task switch leaves residue; the cumulative drag is more than the headline switching time suggests

Deep Work and Shallow Work

Cal Newport's framework distinguishes between "deep work" - cognitively demanding work performed in a state of distraction-free focus that pushes cognitive capacity to its limit - and "shallow work" - non-cognitively-demanding logistical tasks that are easy to do while distracted. It is worth naming what kind of thing this is: a practitioner's framework built on top of the task-switching literature, not itself a research finding. The underlying switching costs are well measured. The claim that most value in knowledge work comes from the deep category is an argument, and a plausible one, rather than something anyone has quantified.

Why this distinction matters. Deep work requires the kind of focused attention that the modern environment systematically prevents. Email, Slack, instant messaging, meetings, and the steady drip of notifications break attention before deep work can be sustained. Most professionals report that the best work they have ever done was during periods when they had unusual amounts of uninterrupted time. The opportunity cost of not protecting time for deep work is therefore not just personal preference; it is the cost of doing less of the work that creates the most value.

What protected attention actually does. Sustained focus on a complex task allows the brain to load enough context to make progress that fragmented attention cannot. Solving difficult technical problems, writing well, designing something complex, doing serious analysis, or learning a new skill all require this. Whether the professionals who protect focus time systematically outperform equally able peers who do not is the claim you would most want tested, and it has not been - the people who arrange their working lives that way differ from those who cannot in ways that would confound any comparison.

How to actually protect it. Time-blocking specific hours of the week with explicit deep-work commitments. Working from environments designed for focus (libraries, dedicated workspaces without notifications). Phone in another room, not just on silent. No email or messaging open during the block. Communicating boundaries to colleagues so they expect delayed response during focus time. None of this is unusual or extreme; collectively it describes what the people who get the most done routinely arrange.

The hours that produce the most value are usually the ones least available in modern work environments

Children and Attention

Children's attention systems develop through use. The capacity for sustained focus on a single activity - reading a book, building with blocks, drawing, having an extended conversation - is partly innate and substantially shaped by environment. Children whose environments support sustained focus develop different attention profiles from children whose environments demand constant rapid switching. The research on this is substantial; what it implies for screen-rich childhoods is less settled than the volume of it suggests.

What the research shows, and what it does not. Heavy screen exposure in early childhood is associated with weaker attention, language development, and self-regulation by school age. The association is consistent across studies; the causal reading is not secure, because the research is overwhelmingly observational and the families whose toddlers watch the most screens differ in many other ways that also affect development. Effect sizes in the better-controlled studies are small. Two mechanisms are proposed - substitution, where screen time displaces the conversation, reading aloud, and free play that build sustained attention, and a direct effect from the rapid pacing of child-targeted content - and the substitution account is the better supported of the two. The American Academy of Pediatrics has recommended avoiding screen media other than video-chat under 18 months and limiting use through age 5, though its more recent guidance leans toward family media planning rather than fixed hour limits, partly in recognition of how mixed the underlying evidence is.

For older children and adolescents. The questions shift from screen time generally to specific platforms and content, and the evidence gets weaker rather than stronger. The sleep effect is the best established: late-evening phone use displaces sleep, and adolescent sleep loss has well-documented consequences. Beyond that, whether heavy social-media use causes declines in attention and mental health, or whether struggling adolescents use more of it, is genuinely contested among researchers who have looked at the same datasets - the Mental Health piece lays out both cases at length rather than settling it here. What is not in dispute is that adolescence in 2026 is a different developmental experience than adolescence in 2010; what that difference has done is the argument.

What helps build attention. Reading aloud to children from early ages. Having conversations rather than narrating around children. Free play (especially outdoor, unstructured, unsupervised within reason). Making time and space for boredom (boredom is when sustained focus most easily develops). Limiting screen time. Avoiding the always-on stimulation that modern environments produce. None of this requires unusual parenting; it requires attending to what specific environments support what specific developments.

The environments that build attention in children look unusually calm by modern standards

Boredom and the Default Mode Network

One of the more counter-intuitive findings of modern neuroscience is that the brain is not idle when you are not focused on anything specific. Marcus Raichle's lab identified a network of brain regions, now called the "default mode network," that is active during what looks like rest - when you are walking without listening to anything, taking a shower, doing dishes without a podcast, sitting on a porch. Exactly what this network does is more contested than popular accounts suggest - the most-supported functions are mind-wandering, self-referential thinking, and social cognition, with the relationship to "creativity" and "insight" still being actively researched - but it is well-established that the brain remains engaged in substantial processing during apparent rest, including consolidating memories and integrating recent experiences.

What this means for attention. The brain needs periods of un-stimulated time to do this background processing. The modern environment fills almost every spare moment with stimulation - phone in line at the coffee shop, podcast during the commute, video while doing dishes, scrolling in bed before sleep. Each of these is individually small; collectively they have eliminated most of the unstimulated time that the brain previously had. The default-mode-network research suggests this matters - the moments of insight, creative connection, and emotional processing that previously happened during walks and showers and quiet moments now have less space to occur.

What this means for boredom. Boredom feels uncomfortable, and the popular claim is that it is the precondition for sustained focus and creative insight. It is worth being straight about how much of that is established. What is well supported: the brain does substantial background processing during apparent rest, and small laboratory studies find that undemanding, boring tasks improve performance on subsequent idea-generation tests. What is not established: that filling spare moments with a phone measurably reduces creativity or attention over a life, which is a much larger claim and one nobody has run the experiment for. Reclaiming some boredom - putting the phone down during routine activities, walks without a podcast, sitting still without input - is a reasonable bet on that evidence rather than a demonstrated intervention, and it is worth trying on those terms.

The brain needs unstimulated time to do the work that conscious focus does not

Working With Attention

Some practical implications, drawn from the research, applied to ordinary life.

Phone hygiene. The single highest-leverage intervention for most adults. Notifications off by default, with explicit exceptions only for things that genuinely need immediate response. Phone in another room while doing focused work or having dinner. No phone in the bedroom at night. Specific apps (social media, news, anything algorithmic) deleted from the phone and accessed only on a computer to add friction. None of this is anti-technology; it is reclaiming the attention that the technology, by default, captures more aggressively than is in your interest.

Single-tasking. When doing anything requiring genuine cognitive engagement - writing, complex analysis, difficult conversations, reading something serious, learning something new - close everything else. One task at a time. Time-block specific hours for it. The output quality is dramatically better and the experience is more satisfying than fragmented multitasking, which is much of what most people end up doing by default.

Protecting time for deep work. If you do knowledge work for a living, schedule explicit blocks of focused time, defend them against meetings and interruptions, and use them for the highest-leverage activities. The claim sometimes made - that two protected hours beat eight fragmented ones - is a rule of thumb rather than a measured result, and the honest version is narrower: task-switching costs are well documented, so fewer switches should mean more output on demanding work. What makes this hard is not discipline but accounting: the cost of being unavailable is immediate, visible, and lands on a named colleague, while the cost of always being available is diffuse and lands on nobody in particular.

Reclaiming boredom. Allow some moments of un-stimulated time in your day. The walk without a podcast. The line without checking the phone. The shower without anything playing. The drive with the radio off. Each of these gives the default-mode network space to do the background processing it needs. The discomfort of boredom is the price of admission to the cognitive states that produce insight, creativity, and integrated thinking.

Designing for attention rather than against it. Most modern environments are designed against attention - open offices with constant interruption, smartphones with always-on notifications, workflows that reward responsiveness over depth. Where you have control - your home, your workspace, your devices, your routines - design for attention. Quiet spaces. Defined work blocks. Tools that minimise notifications. Environmental cues that support focus rather than distraction. The cumulative effect across a working life is enormous.

Being present. The simpler version of all the above is to be present where you actually are. With the people you are with. In the activity you are doing. The amount of life that gets lost to divided attention - dinners barely registered because of phone use, conversations half-tracked, walks taken while elsewhere mentally - is among the largest invisible costs of modern living. Reclaiming presence is partly about specific behavioural changes; it is also about valuing the experience you are actually having over the alternative experiences that other channels are advertising. The trade-off is real; choosing presence over more efficient consumption of stimulation is a small daily decision that compounds dramatically over decades.

Most of what looks like discipline is the residue of environment design done weeks ago

Attention is the slice of reality you actually live in. The modern environment is engineered to capture more of it than is in your interest. The cognitive science on what attention is, how it fragments, and what protects it is substantial and largely unused by the people who would benefit most from it. The practical interventions are not glamorous - notifications off, phone in another room, single-tasking, time-blocking, occasional boredom - and they compound across decades into a dramatically different life than the one default attention-management produces. The choice of where attention goes is, finally, the choice of what life you live.