Attention Span Crisis: What Science Actually Says

The Goldfish Myth and What Was Actually Measured

No, humans don't have an 8-second attention span, and no credible study has ever measured a single "average attention span" in seconds for the whole population. The best-documented number is narrower: in a 2016 study of 40 information workers at one company, the average stretch spent on a single computer screen before switching was 47 seconds, with a median of 40. That's a workplace screen-switching measurement, not a biological limit, and it's now a decade old.

Start with the myth, because it's worth seeing how thin it is. You've met it everywhere: "the average human attention span is now 8 seconds, less than a goldfish." It comes from a single slide in a 52-page 2015 advertising report by Microsoft Canada's Consumer Insights team. The slide reads "12 seconds: The average human attention span in 2000. 8 seconds: The average human attention span in 2013. 9 seconds: The average attention span of a goldfish," and cites a website called Statistic Brain.

BBC journalist Simon Maybin chased it down for More or Less in March 2017. He contacted the sources Statistic Brain listed, the National Center for Biotechnology Information at the US National Library of Medicine and the Associated Press, and neither could find any record of the research. Statistic Brain never replied to him. Gemma Briggs, a psychology lecturer at the Open University who studies attention, told him plainly: "I don't think that's true at all." Her reason: "I don't think that that's something that psychologists or people interested in attention would try and measure and quantify in that way." Attention, she said, "is very much task-dependent."

Here's the part almost nobody mentions. The Microsoft report's own foreword, written by consumer insights lead Alyson Gausby, is headed "Think digital is killing attention spans? Think again," and goes on: "I never would have guessed that tech savvy consumers are actually getting better at processing information and encoding that information to memory." A document arguing that digital isn't shrinking attention has spent ten years being cited as proof that it is.

The goldfish half is wrong too. Felicity Huntingford, who has studied fish behavior for roughly half a century, told the BBC that goldfish "have become a model system for studying the process of learning and the process of memory formation, exactly because they have a memory and because they learn." The animal chosen as shorthand for forgetfulness is a laboratory model for memory.

The same pattern shows up in the classroom version of the claim. The widely repeated idea that student attention collapses 10 to 15 minutes into a lecture was traced back by Neil Bradbury in a 2016 review, and he found that "the most often cited source for a rapid decline in student attention during a lecture barely discusses student attention at all." The available primary data, he concluded, don't support a 10 to 15 minute limit.

So what happens when somebody actually tries to measure attention span as a duration? Adam Gazzaley's group at UCSF did it in 2023, defining "A-span" as the longest continuous stretch a person stays in the zone during a sustained-attention test, across 262 people aged 7 to 85. Children averaged about 30 seconds. Young adults averaged about 76. Older adults, about 67. The figure that exists is task-specific, it's nothing like 8 seconds, and it doesn't correlate with the reaction-time measures the rest of the field relies on.

---

Attention Span Statistics: The Numbers That Hold Up

The most defensible attention span statistics are these: 47 seconds on one computer screen before switching (n = 40, 2016), 11 minutes 4 seconds inside a single work task (n = 24, 2005), and 46.9% of waking hours spent thinking about something other than the task at hand (n = 2,250, 2010). Each has a traceable source, a real sample, and a method you can state in one line. Here they are in full.

And here are the ones to stop repeating. Each of these is in wide circulation, and each falls apart on contact with its source.

Two notes on reading the first of those tables. The 275-pings figure describes the most heavily pinged fifth of workers, which Microsoft says in its own footnote and almost every citation drops. And Killingsworth and Gilbert's 46.9% is from 2010, before TikTok existed; it's a fact about minds, not about phones.

---

What the Phone on Your Desk Actually Does

The most-cited claim in this whole field is that a phone sitting silently on your desk drains your cognitive capacity. It's a great story. It's also the claim that has aged worst.

The original is real and carefully done: Adrian Ward, Kristen Duke, Ayelet Gneezy and Maarten Bos, publishing in the Journal of the Association for Consumer Research in 2017 under the title "Brain Drain." Participants did cognitive tests with their phone on the desk, in a bag, or in another room. Performance on working memory and fluid reasoning got better as the phone got further away.

Two things usually get misreported. First, the study found no effect on sustained attention, none at all: the authors report no effects of phone location on either sustained-attention measure. That's not a footnote, it's the point of the paper, which argues that capacity drops even while attention holds steady. Second, the phones weren't powered off in the main condition; they were face down and silent.

Then the replications arrived. A pre-registered direct replication by Ruiz Pardo and Minda in Acta Psychologica (2022), with a larger sample than the original, found no difference between phone-location conditions on either task. A 2024 meta-analysis by Hartanto and colleagues pooling 166 effect sizes across 33 studies puts the overall effect at d = −0.02, which is indistinguishable from nothing. A separate 2024 meta-analysis by Douglas Parry (56 studies, 7,093 participants) found exactly one significant pooled effect across all cognitive functions, on working memory capacity, and null results everywhere else.

So keep your phone in another room. Just not for this reason. Put it there because it makes an interruption physically impossible, and interruptions are the part of this story that survives scrutiny.

They're relentless, though the honest version is more specific than the headline. Microsoft's 2025 analysis of Microsoft 365 telemetry found that among the most heavily pinged fifth of workers, a meeting invite, email or chat lands every two minutes across an eight-hour day, totalling about 275 across a full 24 hours. That's the top quintile, not the average worker, and it counts pings received rather than attention actually broken. In the same 2025 Work Trend Index survey of 31,000 knowledge workers across 31 markets, eight in ten said they lacked the time or energy to do their work.

The recovery cost is real but smaller than folklore says. Gloria Mark, Victor González and Justin Harris shadowed 24 information workers for more than 700 hours and found that when an interrupted task got picked up again the same day, 25 minutes and 26 seconds had passed. But that's elapsed time with an average of 2.26 other work activities inside it, and the authors themselves describe it as "a relatively short amount of time." The famous "23 minutes 15 seconds" appears nowhere in that paper; it comes from a 2008 Fast Company article.

What does compound is attention residue, a term Sophie Leroy introduced in a 2009 paper in Organizational Behavior and Human Decision Processes. Switch from Task A to Task B and part of your attention stays stuck on A, degrading performance on B for a while afterwards. Stack enough switches and you spend the day at partial strength.

---

The Short-Form Video Effect

Short-form video eats a serious amount of the day: TikTok's Android users average 1 hour 37 minutes in the app, about 14% more than YouTube (DataReportal, 2026). The research linking heavy use to weaker attention is genuinely substantial. It's also almost entirely correlational, and the experiments complicate the story people usually tell.

The strongest synthesis is a 2025 meta-analysis by Nguyen and colleagues in Psychological Bulletin, pooling 71 studies and 98,299 participants. Short-form video use correlated negatively with cognition overall (r = −0.34), with attention (r = −0.38), and with inhibitory control (r = −0.41). Mental health showed a weaker link (r = −0.21), and body image and self-esteem showed essentially none. Those last two nulls matter: this isn't a blanket "short video is bad for you" result, it's specific.

Correlation is the ceiling there, so the experiments carry extra weight. Wei and colleagues published three experiments in Communications Psychology in May 2026 (n = 180, 185 and 59) in which people learning from short-form video had significantly worse immediate memory and forgot faster. The brain imaging is worth quoting carefully, because it doesn't match the usual narrative: synchrony dropped in parietal, precuneus and middle occipital regions, and rose in temporal and frontal ones. If you've read that short video simply switches your frontal lobe off, the best experimental evidence says otherwise.

And the mechanism everyone reaches for, that rapid cuts train your brain to demand novelty, doesn't hold up on its own. A 2025 meta-analysis by Hinten, Scarf and Imuta in Developmental Science, covering 19 studies and 1,431 children, found that the pace of media had no significant effect on children's attention or executive function (d = −0.123, 95% CI −0.331 to 0.086). What did predict an effect was fantastical content, not fast editing.

There's also a consistent finding that what matters is the relationship, not the clock. A 2026 review in JMIR pooling 58 studies and 96,676 participants found that problematic short-video use correlated with lower wellbeing (r = −0.25), while ordinary daily viewing time showed no significant association with negative affect.

One neuroscience result is worth reporting precisely because it's so often stretched. Yan and colleagues (Frontiers in Human Neuroscience, 2024, n = 48, 45 analysed) found that heavier short-video addiction scores went with weaker prefrontal theta power during an attention task, r = −0.395. The same paper explicitly did not find it in resting-state data. So it's a difference in how people's brains handled a task, not a permanent trait you can point at.

For a practical read, this is what the evidence supports: heavy, compulsive short-video use goes with worse attention, the causal direction isn't settled, and swapping some of it for material you actually process is a reasonable bet. Turning a video into something you read, mark up and think about is one way to do that. YouTube Summary by Glasp gives you the transcript and the key points, so a video becomes text you can work through rather than footage that works through you.

---

Attention Span by Generation: What Has and Hasn't Been Measured

Nobody has ever measured attention span in seconds by generation. Not for Gen Alpha, Gen Z, Millennials, Gen X or Boomers. Any table you've seen assigning each cohort a number of seconds is invented. What has been done, once, is a proper cross-temporal comparison of an actual attention test, and it found the opposite of the popular story: adults have been getting better.

That study is a 2024 cross-temporal meta-analysis by Andrzejewski, Zeilinger and Pietschnig in Personality and Individual Differences. It pooled 287 independent samples, 21,291 people across 32 countries, tested on the d2 Test of Attention between 1990 and 2021. The finding, in the authors' words, was "moderate generational test score gains in concentration performance in adults, but not children," alongside "a substantial increase in overall errors and processing speed in children." Read that carefully: children didn't get worse at concentrating. They got faster and less accurate, which is a change in how they approach a test, not a loss of capacity.

So what do we actually know about each cohort?

One caveat before you read it: Nielsen reports age bands, not generations, so in 2018 the 50 to 64 band still held older Gen X alongside Boomers. Two things tend to surprise people anyway.

First, the screen-time ranking flips depending on what you count. If you count phones only, younger cohorts are on top. If you count total media time including television and radio, older cohorts win by hours: Nielsen's 2018 panel measurement put 50 to 64 year olds at nearly 13 hours a day against under 9 for 18 to 34 year olds. Both are true. Neither is an attention measurement.

Second, older adults don't do worse on sustained attention in the lab. A 2021 meta-analysis by Vallesi and colleagues covering 12 studies and 1,522 participants found older adults were slower on go trials but significantly more accurate on no-go trials (g = 0.59). Mind-wandering runs the same way: Jackson and Balota (2012) found older adults reported less of it, and Krawietz and colleagues found that difference disappeared once you controlled for how interested people were in the passage. What declines with age is processing speed and working memory, not the ability to stay on task.

It's also worth naming the evidence that gets substituted for attention data. The most-cited proof that "cognitive ability is falling across cohorts" is the Flynn effect reversal, and the flagship study is Bratsberg and Rogeberg's 2018 PNAS paper on Norwegian conscripts. It's excellent work on a huge sample, and the scores really do turn around with the 1975 birth cohort. But the battery consists of arithmetic, word similarities and figures. There is no attention test in it, no vigilance test, no reaction-time test. The same goes for NAEP and PISA, which measure what students learned, not how long they can hold focus. And NAEP's 9-year-olds gained 4 points in both reading and maths in 2025 against 2022, while 13-year-olds showed no significant change.

Teacher surveys get quoted as measurement too. Pew found in 2024 that 72% of US high school teachers call cellphone distraction a major problem in their classroom, which is a real and useful finding about classrooms. In the same research program, 45% of teens said smartphones make school easier against 23% who said harder, and 70% said phones bring more benefits than harms for people their age. Adults and teenagers are describing the same phones and disagreeing about the sign.

One last substitution to watch: rising ADHD diagnoses. Diagnoses have genuinely climbed, from 8.6% of US children in 2016 to 10.5% in 2023. But the analysis that held the diagnostic method constant across three decades, Polanczyk and colleagues in 2014, concluded that "there has been no evidence to suggest an increase in the number of children in the community who meet criteria for ADHD when standardized diagnostic procedures are followed." More children are being identified. That isn't the same as more children having the condition.

The honest summary for any generation, including yours: your environment changed enormously, your measured capacity for sustained attention did not, and the one time somebody checked properly, adults had improved.

---

The Real Cost of Distraction

Fragmented attention isn't just annoying. It carries measurable costs across work, learning, and wellbeing.

Productivity and workload. When pings arrive every couple of minutes for the most-connected workers, the time left for deep work shrinks. There's a wrinkle worth knowing, though: Mark, Gudith and Klocke's 2008 experiment found that interrupted work was actually completed faster, with no drop in quality, but at the cost of significantly higher stress, frustration and time pressure. People compensate. It just hurts.

Reading, and the shape of its decline. A 2025 study in iScience by Bone, Bu, Sonke and Fancourt analysed 20 years of the American Time Use Survey (236,270 respondents, 2003 to 2023) and found the share of Americans reading for pleasure on a given day fell from a peak of 28% in 2004 to 16% in 2023, a relative decline of about 3% a year. The detail that changes the story: among people who did read, daily reading time went up, from 1 hour 23 minutes to 1 hour 37 minutes. This is a participation collapse, not an attention collapse. Fewer people read at all, and the ones who do read more than they used to. Bureau of Labor Statistics figures published since put the share at 16.1% in both 2024 and 2025, so the slide has flattened. Among UK children it has turned around: the National Literacy Trust's 2026 survey of 125,375 children found reading enjoyment up to 36.1% from 32.7%, and daily reading up to 20.3% from 18.7%, the first rise in five years. Different country, different ages, different measure, so it doesn't rebut the US figures. It does show the trend isn't a law of nature.

Learning quality. Sana, Weston and Cepeda (2013) found that students who multitasked on laptops during a simulated lecture scored lower on comprehension, and so did students seated where they could see those screens. It was a controlled simulation rather than a real class, and the multitasking was assigned by the experimenters, but the peer effect is a striking result and it replicates the intuition every teacher has.

The multitasking illusion. Sanbonmatsu and colleagues (2013) found that people who multitask most tend to rate themselves as good at it and perform worse on a working-memory measure, though the correlation was modest (r = −0.19). The "supertasker" finding, that about 2.5% of that sample showed no measurable decrement, comes from a single lab and hasn't been independently replicated in 16 years. Treat it as a curiosity, not a personality type you might belong to.

Mental health. Gloria Mark's research links frequent task-switching to higher self-reported stress, and the relationship plausibly runs both ways: fragmented attention raises stress, stress fragments attention further.

Regulators are now treating this as design, not accident. In July 2026 the European Commission preliminarily found that Instagram and Facebook breach the Digital Services Act through addictive design, naming infinite scroll, autoplay, push notifications and highly personalised recommender systems. Whatever happens to that case, it puts the "these products are engineered to fragment your attention" argument in the hands of an enforcement authority rather than a critic.

---

What Actually Works: Evidence-Based Focus Strategies

Attention responds to training and to environment. Here's what the evidence supports, with the strength of that evidence stated honestly rather than as a five-star rating.

Put the phone in another room, for the right reason. Distance doesn't stop some mysterious cognitive drain, because that effect vanished under replication. It stops the interruption from reaching you at all. That's a smaller claim and a much sturdier one.

Design the environment instead of relying on willpower. Gloria Mark's work consistently shows external cues drive attention more than self-control does. Close unnecessary tabs, work full screen, and batch email and chat into two or three windows instead of checking continuously.

Time-box, but don't fetishise the timer. Francesco Cirillo's Pomodoro Technique works for a lot of people because a fixed boundary makes open-ended focus less daunting. The 25-minute figure is Cirillo's, not a research finding. Define a start, an end, and one task; the interval is yours to tune. If you want the full method, we've written it up in the Pomodoro Technique for studying.

Train attention directly, with realistic expectations. Amishi Jha's Peak Mind popularised a 12-minute daily practice, and the underlying work is real, but two caveats usually get dropped. The 12 minutes is a threshold observed among people already enrolled in an eight-week, 24-hour in-person course, not a standalone dose anyone was randomised to. And the outcome in that study was working memory, not sustained attention; Jha's later sustained-attention work used 30 minutes a day. A 2024 review of 111 randomised trials found genuine but small effects against active control conditions. It works. It isn't magic.

Reduce social media rather than quitting it. The well-known Penn study by Hunt and colleagues (2018) limited participants to 10 minutes per platform per day across three platforms and found reductions in loneliness and depression. Two details are usually lost: that's roughly 30 minutes total, not 30 minutes on one app, and both groups improved on anxiety and fear of missing out, which the authors attribute to the self-monitoring itself. A 2025 pre-registered meta-analysis of abstinence studies found no significant effect on affect or life satisfaction, while reduction studies continue to show a small benefit for depression. Cutting back beats going cold turkey.

---

Active Reading as Attention Training

Most focus advice is subtractive: remove the distractions. That's necessary and incomplete. You also have to give attention something to hold onto, and this is where a lot of study advice overreaches, so it's worth being exact.

Here's the uncomfortable evidence first, because Glasp is a highlighting product and you should know we've read it. In their landmark 2013 review in Psychological Science in the Public Interest, John Dunlosky and colleagues assessed ten study techniques and gave highlighting a low utility rating. Their verdict was blunt: highlighting "does little to boost performance," and may even "hurt performance on higher-level tasks that require inference making." Summarizing and rereading landed in the same low-utility group.

The two techniques that earned a high utility rating were practice testing and distributed practice. That's the finding to build on, and it reframes what highlighting is for. A highlight is not a study method. It's a capture step, and its value is entirely in what you do with it afterwards. Marking a passage and never returning to it does roughly nothing. Marking a passage, then using it as the prompt for a retrieval attempt a few days later, hooks you into the technique with the best evidence in the whole review.

Dunlosky left a door open on this, too. Highlighting "may help when students have the knowledge needed to highlight more effectively, or when texts are difficult," and he called for research on teaching people to highlight well. The failure mode is indiscriminate marking, not marking as such.

There's a second, separate claim worth keeping clean: that annotation holds attention in place while you read. That's about engagement rather than memory, and Dunlosky's review didn't test it. The mind-wandering research is suggestive here. Reichle, Reineberg and Schooler (2010) documented "mindless reading," where the eyes keep moving across the page while the mind is elsewhere, in an intensive four-participant eye-tracking study in which readers still answered 81% of comprehension questions correctly (chance was 25%), and Smallwood, McSpadden and Schooler (2008) found that readers who had zoned out identified the culprit in a detective story 30% of the time against 61% for those who hadn't. Comprehension halves; it doesn't disappear. Anything that forces a decision every few paragraphs ("is this worth marking, and why?") gives a wandering mind something to catch on.

One popular citation doesn't belong here at all, and we used to make this mistake. Mueller and Oppenheimer's 2014 study is about handwriting versus typing during lectures. It says nothing about highlighting or about reading. It also hasn't held up: Morehead, Dunlosky and Rawson (2019) found performance "did not consistently differ between any groups, including a group who did not take notes," and Urry and colleagues (2021, N = 142) found no significant difference at all, with a point estimate slightly favouring laptops (g = −0.13, 95% CI −0.45 to 0.20). A mini meta-analysis across eight studies puts the effect on quiz performance at g = 0.04, which is zero. What does replicate is what ends up in the notes: laptop users write more words and copy more verbatim.

So here's the defensible workflow. Read actively and mark deliberately, keeping your highlights few enough that each one meant something. Then close the loop. Glasp's web highlighter saves what you marked into a searchable library across everything you read, including Kindle highlights imported from your books, which is what makes the return visit cheap enough to actually happen. Come back to those passages later and try to recall the argument before you reread it. That's active recall running on material you already chose, and it's the part the evidence actually backs.

---

Building a Deep Focus Practice

You wouldn't run a marathon without training. Don't expect to sustain multi-hour deep work if your day currently runs in 47-second screen cycles. Build gradually.

Weeks 1 and 2: establish a baseline. Start with two 25-minute focused sessions a day (the length is arbitrary, the boundary is the point). Phone in another room. Notifications off. One task. Track how often you feel the urge to switch, without judging it. That count usually drops on its own by the second week, mostly because you start noticing it.

Weeks 3 and 4: extend duration. Move to 35-minute sessions and make one of them active reading. Use Glasp to highlight and annotate one long-form article a day. Longer sessions plus active engagement build stamina faster than either alone.

Weeks 5 to 8: deepen the practice. Aim for 50-minute blocks with 10-minute breaks, scheduled on your calendar like meetings. On the ceiling, be realistic: Cal Newport, citing Anders Ericsson, notes that novices manage roughly an hour a day of intense concentration while experts can reach as many as four hours, "but rarely more." Most people aren't at the expert end, and that's fine. The gap between one hour and zero is the one that matters.

The daily attention recovery framework:

Create friction for distraction. Small amounts of friction cut impulsive behavior sharply. Phone in a drawer rather than on the desk. Website blockers during focus periods. Log out of social accounts so getting back in takes effort. Disable non-essential notifications permanently. You don't need to win a willpower battle, just make the distraction slightly harder to reach.

Take a weekly break from the feed. Building on the reduction research, a regular low-social-media day lets your attention recalibrate away from rapid-switching mode. One deliberate day a week is enough to notice.

Consider applying slow reading during your focus blocks. Reading deliberately, pausing to reflect, and rereading hard passages isn't inefficiency. It's the deep processing that skimming has crowded out.

---

Frequently Asked Questions

Do humans really have an 8-second attention span?

No. The 8-second figure, usually paired with "less than a goldfish," is fabricated. It appears on one slide of a 2015 Microsoft Canada advertising report that cited a website called Statistic Brain, and when the BBC tracked down the institutions Statistic Brain listed in 2017, none could produce any research behind it. The report's own foreword argues that digital media is not shrinking attention. Goldfish, meanwhile, are a standard laboratory model for studying memory.

What is the average attention span in 2026?

There isn't a credible single number, and anyone quoting one to the second is repeating the goldfish myth. The closest measured figures are task-specific. Average time on one computer screen before switching was 47 seconds in a 2016 workplace study of 40 people. When researchers tried to measure a continuous "attention span" on a sustained-attention test in 2023, children averaged about 30 seconds, young adults about 76, and older adults about 67.

What is Gen X's attention span?

Nobody has measured it, and neither has anyone measured it for any other generation. No study has given the same attention task to different birth cohorts at the same age. What's known about Gen X specifically: in Nielsen's 2018 panel measurement, 35 to 49 year olds consumed about 11 hours 9 minutes of total media a day, considerably more than 18 to 34 year olds at 8 hours 45 minutes. No laboratory vigilance study isolates this age band either. The meta-analytic age effect that does exist, slower responses paired with fewer false alarms, compares younger adults with older adults. Slower and more accurate is the ageing pattern, not a Gen X result.

Do younger generations have shorter attention spans?

There's no evidence that they do. Attention has been compared across birth cohorts exactly once with a real attention test: a 2024 meta-analysis of 287 samples (N = 21,291, 32 countries, 1990 to 2021) using the d2 Test of Attention. It found moderate gains in concentration performance among adults and no decline in children, though children did become faster and made substantially more errors, a shift in the speed-accuracy trade-off rather than a measured loss of concentration. The evidence usually offered instead, the Flynn effect reversal and falling test scores, comes from batteries of arithmetic, vocabulary and reasoning items that contain no attention test at all.

What is Gen Z's attention span?

Nobody has measured that either, and the figures in circulation are invented. What has been measured is exposure: US 13 to 18 year olds averaged 8 hours 39 minutes of entertainment screen media a day against 34 minutes of reading (Common Sense Media, 2021), and 36% of teens told Pew in December 2025 they use one of five major platforms almost constantly. The only cross-cohort comparison using a real attention test found gains in adults and no gains in children, who became faster and less accurate rather than less able to concentrate.

Can you actually increase your attention span, or is it fixed?

You can improve it, within limits. Environmental changes work immediately: phone in another room, fewer notification sources, one task at a time. Mindfulness training produces real but small gains against active control conditions across 111 randomised trials. Reducing social media has a modest effect on mood, and reduction outperforms total abstinence. Think of it less like a muscle you can grow indefinitely and more like sleep hygiene: mostly about removing what's degrading it.

Is short-form video really shrinking my attention span?

Heavy use is consistently associated with weaker attention, with a pooled correlation of r = −0.38 across 71 studies and 98,299 people. But the evidence is correlational, so the direction isn't settled, and the usual mechanism story is shakier than you'd expect: a 2025 meta-analysis of 19 experiments found media pace had no significant effect on children's attention. What does predict problems is compulsive use rather than viewing time. Experiments do show worse memory and faster forgetting when people learn from short video rather than text.

Does multitasking actually work for some people?

Almost certainly not for you, and the famous exception is weaker than it sounds. Sanbonmatsu and colleagues (2013) found heavy media multitaskers scored worse on a working-memory measure, though the correlation was modest. The "supertasker" result, that about 2.5% of one sample showed no decrement, comes from a single laboratory and has not been independently replicated in the 16 years since. What feels like effective multitasking is rapid task-switching with attention residue accumulating at each switch.

How does your information diet affect your attention span?

They're closely connected. High volumes of short-form, algorithmically selected content go with expecting a new reward every few seconds, which can make a 3,000-word article feel harder than it is, though that link is correlational rather than demonstrated training. A deliberate information diet, weighted toward long-form reading and curated rather than algorithmic sources, keeps that expectation in check. Glasp's community feed is one way to do the curation with human judgment, since you're seeing what thoughtful readers chose to highlight rather than what an engagement model surfaced.

---

Conclusion: Reclaim Your Attention, Reclaim Your Mind

The attention crisis is real, but it's a smaller and stranger thing than the headlines say.

What's real is the environment. Pings every couple of minutes for the most-connected workers. Feeds engineered so well that a regulator has started calling the design addictive. Reading on a given day getting rarer still, down to 16% of Americans picking up something for pleasure. Those are documented, and they're worth taking seriously.

What isn't real is the idea that your brain has been damaged, or that you've been issued a shorter attention span than the generation before you. The one time researchers compared attention across cohorts with an actual attention test, adults had improved. The 8-second statistic came from an advertisement. The 23-minute recovery figure came from a magazine interview. The silent phone on your desk turned out not to drain anything beyond, at most, a small working-memory effect. Strip out the folklore and what's left is a straightforward environmental problem, which is much better news than a neurological one, because environments can be redesigned.

Start small and start tomorrow. One 25-minute phone-free block in the morning. One long-form article read properly, with a few deliberate highlights in Glasp and a return visit later in the week to see what you remember. One evening without the feed. None of that is dramatic. All of it compounds.

Your attention is the thing every other capability runs on. It's worth defending, and the evidence says it's in better shape than you've been told.