Originally published in December 2019 and comprehensively updated/edited in June 2026.
How information overload, social media, smartphones, and the attention economy pushed human cognition toward its limits in the years before 2020 — and what the science says about why.
There was something unusual happening to the human mind in the years leading up to 2020. Not a dramatic rupture — no single moment of before and after — but a slow, accumulating pressure. People were processing more information than any generation in history, making more decisions, managing more relationships, and doing all of it at a pace that the architecture of the human brain had simply never been asked to sustain. The tools were extraordinary. The cognitive cost was significant. And most people had no framework for understanding what was happening to them.
This is an examination of that period — the psychological, neurological, social, and technological forces that converged in the late 2010s to push humanity toward what might reasonably be called a cognitive threshold: a point at which the demands placed on human attention, memory, and emotional regulation began to exceed what our biology was designed to handle. Understanding that threshold, and how it was crossed, matters more now than it did then.
The Architecture of Attention Was Never Built for This
To understand what was happening to cognition around 2019, it helps to understand what attention actually is and what it isn’t. Popular culture tends to treat attention like a spotlight — something you either have pointed at the right thing or you don’t. The neuroscience is considerably more interesting.
Attention is better understood as a limited resource managed by overlapping networks in the brain, particularly the prefrontal cortex, which handles executive function, and the default mode network, which activates during rest and internally directed thought. These systems have a metabolic cost. They run on glucose and oxygen. They fatigue. And crucially, they were shaped by an evolutionary history in which the amount of information competing for them was orders of magnitude smaller than what existed in 2019.
Research by psychologists Roy Baumeister and colleagues, beginning in the late 1990s and continuing through the 2010s, popularized the concept of “ego depletion” — the idea that self-control and deliberate decision-making draw on a shared pool of mental resources that can be exhausted. While some aspects of that original model have been debated and refined since, the underlying observation held weight: sustained cognitive effort is not free, and humans accumulate deficits across the day. By the late 2010s, researchers increasingly argued that modern life required vastly more daily decisions than previous generations, contributing to growing interest in decision fatigue and cognitive overload as serious public health concerns.
The notification economy had created a world of near-constant interruption. Gloria Mark, a professor at the University of California Irvine who has studied workplace attention for decades, found in research published during this period that it took an average of 23 minutes for a person to fully return to a task after being interrupted. Meanwhile, office workers were being interrupted — or self-interrupting — every few minutes. The math was punishing.
This bred what technology writer Linda Stone had termed “continuous partial attention” — a state of remaining perpetually alert to incoming information without fully committing attention to any single task. Unlike focused concentration or genuine rest, continuous partial attention kept people in a kind of low-grade mental readiness that was exhausting precisely because it was never fully resolved. It was not multitasking; it was the cognitive equivalent of never quite sitting down.
What Smartphones Did to the Brain
The smartphone was not simply a communication device. It was a portal into an attention economy engineered by some of the most sophisticated behavioral scientists and product designers in the world, all optimizing for engagement at a neurological level.
Variable reward schedules — the same mechanism that makes slot machines effective — were baked into social media feeds. A scroll through Instagram or Twitter in 2019 produced an unpredictable mix of rewarding content (a funny video, a message from a friend, a validation through likes) and unrewarding content. That unpredictability is precisely what makes such systems compelling. The dopaminergic circuitry of the brain responds more powerfully to unpredictable rewards than to predictable ones, a principle that behavioral psychologists had understood since B.F. Skinner’s mid-twentieth century experiments, but which had never been deployed at planetary scale until smartphones arrived.
Psychologist Jean Twenge, in her 2017 book iGen, documented a striking inflection point in American adolescent mental health data that occurred around 2012 — the year smartphone ownership among teenagers crossed 50 percent. Depression, anxiety, loneliness, and suicide ideation among teenagers began rising sharply at precisely that moment. The correlation was not proof of causation, and the research community continued debating the strength of the causal link through the end of the decade, but the timing was difficult to dismiss entirely.
By 2019, that cohort of teenagers had grown older. The effects, whatever their precise origin, were visible not just in youth but in adults who had adapted their behavior to mobile technology over the preceding decade. Sleep disruption was pervasive — blue light from screens suppresses melatonin production, and the emotional stimulation of social media kept many people in a low-level arousal state that made winding down physiologically difficult. Matthew Walker, a sleep scientist at UC Berkeley, argued in his 2017 book Why We Sleep that the developed world was in the middle of a sleep deprivation crisis with significant cognitive and health consequences. The timing of that crisis aligned closely with the mass adoption of smartphones.
The Social Complexity Problem
Human beings evolved in small social groups — estimates of ancestral band size cluster around 50 to 150 individuals, a range that anthropologist Robin Dunbar famously formalized in the 1990s. Dunbar’s number, roughly 150, represents the approximate cognitive limit of stable social relationships the human brain can maintain. The neocortex, and particularly its prefrontal regions, appears to have expanded in primates in part to manage social complexity. It is, in a meaningful sense, a social computer.
Social media dismantled Dunbar’s constraints overnight. By 2019, the average Facebook user had 338 friends on the platform — already beyond the theoretical limit — and was passively exposed to the lives, opinions, and emotional states of thousands more through shared content and algorithmic amplification. This was not simply an increase in quantity. It was a fundamental change in the kind of social processing the brain was being asked to do.
Managing relationships requires reading emotional cues, tracking reputations, remembering social histories, and navigating shifting alliances. Doing this with a network of 150 people in a physical community is demanding enough. Doing it with hundreds or thousands of people across a digital network — where context collapse meant that a single post could be seen simultaneously by your boss, your high school acquaintance, your political opponent, and your closest friend — introduced a form of social complexity for which there was no evolutionary precedent.
The psychological phenomenon of social comparison, well documented since Leon Festinger’s social comparison theory in 1954, took on new dimensions in this environment. People were now comparing themselves not to their immediate peer group but to curated highlights from an enormous and algorithmically curated social universe. Research published in journals like Computers in Human Behavior throughout the 2010s repeatedly found associations between passive social media consumption and decreased subjective wellbeing, with envy, inadequacy, and social anxiety featuring prominently.
Information Overload and the Collapse of Filtering
Prior to the internet, information filtered itself through institutional bottlenecks: publishers, editors, broadcasters, librarians. These filters were imperfect and often exclusionary in ways that had real social costs. But they performed a function that most people did not consciously appreciate until those filters were gone: they reduced the cognitive burden of deciding what was worth attention. By 2019, those filters had been largely disintermediated by algorithmic feeds that optimized not for quality or relevance in any deep sense, but for engagement — and engagement, research consistently showed, was most reliably triggered by content that provoked strong emotional reactions, particularly fear and outrage.
The result was a media environment that was, in measurable psychological terms, more stressful than its predecessors. The American Psychological Association’s annual Stress in America survey documented rising stress levels throughout the late 2010s, with news consumption consistently identified as a significant contributor. The term “news anxiety” entered clinical parlance. Therapists reported clients describing a compulsive relationship with news and social media that resembled behavioral addictions in its structure — the inability to stop despite negative consequences, the rationalizations, the relief followed by guilt.
Cognitive scientist Daniel Levitin, in his 2014 book The Organized Mind (widely read through the rest of the decade), described the neuroscience of what he called “the modern epidemic of mental unrest and cognitive overload.” The prefrontal cortex, he explained, is remarkably capable but comes with strict bandwidth limits. When those limits are exceeded, decision quality declines, emotional regulation suffers, and creative thinking becomes harder to sustain.
Polarization as a Cognitive Symptom
One of the most visible social phenomena of the late 2010s was the intensification of political and cultural polarization, particularly in the United States but also across the developed world. The standard explanations focused on economic inequality, demographic change, and political manipulation — all of which were real. But there was a cognitive dimension to polarization that received less attention and may have been equally important.
Under cognitive load — when people are tired, overwhelmed, or processing large amounts of information quickly — human thinking tends to revert to heuristics. Psychologists call this System 1 thinking, a concept Daniel Kahneman formalized in Thinking, Fast and Slow (2011). System 1 is fast, automatic, and emotional. System 2 is slow, deliberate, and analytical. The catch is that System 2 requires precisely the kind of cognitive resources that were being depleted by the attention economy.
The platforms made this worse in a specific way: because outrage and fear generated more engagement than nuance, algorithms surfaced emotionally inflammatory content preferentially. This meant that people in an already-depleted cognitive state were encountering a political information environment specifically engineered to trigger emotional responses rather than deliberate analysis. The chain of causation ran from cognitive overload to heuristic thinking, with algorithmic emotional amplification accelerating both.
Tribal categorization, confirmation bias, and affective polarization — developing intense dislike for the opposite political group — are all associated with System 1 processing under stress. They are also all features of the political landscape of 2019. This does not mean polarization was simply a cognitive malfunction. Real disagreements existed about real issues. But the intensity and intractability of those disagreements suggested that something beyond ideology was operating — something that looked, from a cognitive science perspective, like a population under sustained attentional stress defaulting to its mental shortcuts.
When Cognitive Stress Becomes Physical
The distinction between psychological stress and physical stress is largely artificial. The body does not make it. When the mind experiences sustained stress — regardless of whether the stressor is a predator or an anxious scroll through a news feed — the hypothalamic-pituitary-adrenal (HPA) axis activates, releasing cortisol and adrenaline. Over short periods, these stress hormones are adaptive. Over sustained periods, their effects are damaging.
By the late 2010s, researchers studying chronic stress were documenting effects that extended well beyond mood. Chronically elevated cortisol is associated with impairment of the hippocampus, the brain structure most critical to memory formation and consolidation. It is associated with reductions in prefrontal cortex function — the region most responsible for the executive functioning that modern life demanded most. It is associated with increased amygdala reactivity, meaning that chronically stressed people became more emotionally reactive to perceived threats, which created a feedback loop: stress made people more susceptible to stress-inducing content, which generated more stress.
Epidemiological data from the late 2010s showed rising rates of anxiety disorders, which by 2018 had become the most common mental health condition in the United States, affecting roughly 40 million adults. Depression rates were also climbing. These were not merely individual problems. They represented a measurable shift in the population-level distribution of mental health — one whose timing aligned closely with the decade of smartphone mass adoption.
Early Awareness and the Attention Economy’s Critics
Not everyone was oblivious to what was happening. By the late 2010s, a significant conversation had developed — in academia, in journalism, and within the technology industry itself — about the consequences of the attention economy.
Tristan Harris, a former Google design ethicist, had become one of the most prominent critics from inside the industry, arguing that the major technology platforms had deployed techniques derived from behavioral psychology to maximize engagement in ways that were fundamentally exploitative of human cognitive vulnerabilities. His work contributed to the documentary The Social Dilemma (2020), which brought these concerns to a mainstream audience.
Academic researchers like Sherry Turkle at MIT had been arguing for years, most notably in Reclaiming Conversation (2015), that the shift to device-mediated communication was degrading the quality of human connection and the cognitive habits associated with it — sustained attention, empathy, the tolerance for silence and ambiguity that genuine conversation requires. Cal Newport’s Digital Minimalism (2019) offered a framework for intentional technology use that resonated with a growing audience which felt, without necessarily having the vocabulary for it, that something had gone wrong in their relationship with their devices.
These conversations were meaningful. They suggested that humanity was beginning to develop some self-awareness about the threshold it was approaching. But awareness and systemic change are different things.
What Came Next
The COVID-19 pandemic, which began disrupting global society in early 2020, arrived at a moment when the cognitive systems of hundreds of millions of people were already under sustained pressure. Screen time increased sharply, remote work collapsed the physical boundary between professional and personal life, social isolation amplified the emotional reliance on digital communication, and the information environment became, if anything, more chaotic. The pandemic also introduced new and severe stressors — fear, loss, economic precarity — on top of the existing cognitive load. <blockquote>In retrospect, 2019 looks less like a stable baseline and more like a pressure point.</blockquote> The systems — psychological, social, informational — were already strained. The pandemic did not create the cognitive crisis; it stress-tested and revealed the fragility of systems that had been accumulating strain for years.
The years following the pandemic brought generative AI into everyday life, automating categories of intellectual labor that had previously required human attention, while simultaneously raising new questions about authenticity, information quality, and what human cognition is actually for. The structural forces that had driven humanity toward the cognitive threshold did not reverse — many accelerated. But the cultural and scientific conversation about them matured considerably.
The Broader Significance
What the period leading up to 2020 represents, in the long view, is a moment when the pace of technological change outstripped the pace of human psychological adaptation. This is not a new pattern — every major technological transformation produces a period of dislocation before society finds equilibrium. The printing press, industrialization, electrification, and television all created disruptions that took generations to fully absorb. The digital revolution is no different in kind, only in speed.
The human brain is adaptive. Neuroplasticity — the brain’s capacity to reorganize itself in response to experience — means that the cognitive challenges of one era can, over time, produce cognitive capacities suited to meeting them. Whether the demands of the attention economy eventually produce adaptive capacities, or lasting damage, is a question that will require generations to fully answer.
What can be said with reasonable confidence is that by the late 2010s, humanity was navigating a genuine cognitive transition — a period in which the information environment had changed faster than the institutions, habits, and self-knowledge needed to navigate it. The threshold was not crossed with a single dramatic step. It was approached gradually, largely invisibly, in the aggregate of millions of individual moments of distraction, depletion, and decision fatigue. Understanding that threshold, naming it clearly, is part of the work of navigating whatever comes next.
