Or the art of lucidity in the age of artificial intelligence
“Dave, stop. Stop, will you? Stop, Dave. Will you stop, Dave?” Thus does the supercomputer HAL plead with the relentless astronaut Dave Bowman in one of Stanley Kubrick’s most famous and moving scenes from 2001: A Space Odyssey. Bowman, methodical, disconnects the memory circuits of the failing machine one by one, until he extinguishes its artificial consciousness. “Dave, my mind is going,” HAL murmurs in a final digital breath. “I can feel it. I can feel it.” Nicholas Carr, “Is Google Making Us Stupid?” [1]
There is something more than a simple confrontation between man and machine in this scene. This suspended, almost tender moment seems to speak of the confusion between two intelligences watching each other fade into one another. HAL, though pure logic, implores like a conscious being; Bowman, though human, acts with the coldness of an automaton. Between them, the boundary wavers. Kubrick had sensed this troubling shift: it’s not just about whether the machine can think, but about how far man can delegate his own thinking without dissolving into it.
Now imagine the same scene, transposed to our era. You formulate a complex question, nuanced, uncertain. Before you’ve even finished typing it, an answer appears on the screen: clear, articulate, persuasive. Artificial intelligence, this new deus ex machina, has spoken. It hasn’t reflected, it has calculated. And in this minuscule yet dizzying gap between thought and calculation opens one of the most fascinating and most troubling abysses of our time.
In our frantic quest for speed and efficiency, we have ended up confusing the rapidity of calculation with the accuracy of thought. By wanting to accelerate everything, we have delegated not only our tasks but also the very effort of thinking. Artificial intelligence today accomplishes what we have taught it to desire: going faster, producing more, responding without delay. But this promise of fluidity comes at a price. What we gain in performance, we lose in depth. The long time of reflection, the kind that involves error, hesitation, backtracking, becomes an inconvenience in a world where slowness is experienced as a fault.
In a universe where generative AIs are used daily by 42% of young French people [2], and where, according to a Massachusetts Institute of Technology study, 83% of AI users are unable to remember a passage they just had written by the machine [3], thinking for oneself becomes almost an act of indiscipline. Cognitive delegation has morphed into a cultural reflex. We no longer seek to understand, but to confirm. And this substitution of judgment with ready-made answers gradually transforms the act of thinking into a luxury.
But this luxury is not in vain. It carries within it something subversive. In a society that values speed as a sign of success, slowing down to think is already an act of disobedience. It means refusing the tyranny of the flow, refusing to be a simple relay for what circulates. Resisting here doesn’t mean rejecting technology, but opposing it with another temporality: that of discernment, of distance, of silence. Because in the clamor of machines that speak in our place, the slowness of thought becomes a form of courage. The courage to still inhabit doubt, instead of letting ourselves be carried by evidence.
I propose a journey to the heart of this silent mutation, where acceleration is no longer just a question of rhythm, but a way of thinking about the world. We will explore the necessity of slowing down to understand, of letting doubt breathe, and of restoring to slowness its initiatory value. Because thinking slowly is not a step backwards—it’s a reconquest. A way of regaining footing in a universe where everything conspires to rush us.
We will dive into the intimate mechanisms of our mind, guided by the work of Daniel Kahneman, to understand how technology pushes us toward fast, instinctive, but often deceptive thinking. Then we will see, with sociologist Hartmut Rosa, how our societies have locked themselves into a spiral of acceleration that leaves us paradoxically immobile. Finally, we will discover, through “slow philosophy” and contemplative neuroscience, concrete paths to relearn how to exist outside the flow, to cultivate our critical thinking and our lucidity. Because today, thinking slowly is no longer anachronistic—it may be the boldest act to preserve our humanity.
But before exploring the mechanics of our brain, we must first understand the world in which this brain is trying to survive. Because thought doesn’t float in the abstract: it takes root in a context, it breathes (or suffocates) at the rhythm of its era. If we think too fast today, it’s first and foremost because everything around us has accelerated. To grasp how technology diverts our attention and our judgment, we must first take the measure of the force that carries us away: this acceleration that has crept into every corner of our lives, until it has become invisible because it has become our normality.
Living like a hamster running on its wheel
To understand why slowness has become a form of resistance, we must first take the measure of the force opposing it: an acceleration that has crept into every corner of our lives. We turn in the wheel, like those hamsters running endlessly, simply to avoid being ejected from the movement. This diffuse feeling of being “overwhelmed” is not an individual illusion, but the manifestation of a collective disorder. German sociologist Hartmut Rosa, in his masterful work Acceleration: A New Critical Theory of Modernity, dissected this phenomenon with almost clinical precision: a structural mutation of modernity. We run ever faster, not to move forward, but to avoid falling back.
Rosa distinguishes three main drivers of this endless race, three dynamics that drive each other in a dizzying spiral [4].
1. Technological acceleration. This is the most visible, the one that jumps out at you. In just a few generations, we’ve gone from horses to rockets, from handwritten letters to instant messages. Technology has compressed time, erased distances, and yet it has never saturated our days more. The paradox is cruel: what was supposed to free us from time has colonized it. The smartphone, marvel of connectivity, is also an invisible leash. Work has no walls anymore—it slips into our pockets and vibrates in our hands. The expression “I’ll get back to you between meetings” has become a linguistic tic, the banal trace of a fragmented time where every minute must be profitable. As a Medium article on Rosa summarizes, “there is always one more email to send or a project to finish by tomorrow” [5].
Social networks embody the ultimate stage of this acceleration. A piece of information (whether accurate or false) now circles the globe in seconds. This instantaneousness creates an unprecedented pressure: that of permanent immediacy. You must react quickly, comment quickly, share quickly, or risk being swept away by the next wave. The flow of information that passes through us each day is so dense that it short-circuits our capacity for discernment. Designed to process the world in sequences, our brain finds itself plunged into a discontinuous torrent of images and data. We jump from one subject to another without digesting, without understanding, without connecting. This incessant movement, far from expanding our consciousness, exhausts it. We live in a state of cognitive alert where everything is urgent and nothing is important.
2. The acceleration of social change. Our habits, our opinions, our jobs, our relationships—everything seems to wear out at a new speed. The fast fashion industry, with its micro-seasons that succeed each other relentlessly, is merely a caricature of this. What’s at stake is deeper, more diffuse, and touches the very structures of our existence. Skills become obsolete before being fully stabilized, professional trajectories fragment into successive projects, and collective belongings are recomposed at the rhythm of platforms and ephemeral communities. Even large companies struggle to promise lasting stability to their employees, a sign of a horizon that has become shifting [5]. In Rosa’s terms, this volatility is not an accident but the expression of a modernity where permanent transformation becomes the structural norm [4].
This movement doesn’t only affect the economy of work. It reconfigures sociabilities and reference points. Bonds are tied and untied along short temporalities, lifestyles reinvent themselves in waves, norms circulate faster than they take root. We change tools, networks, cultural references with apparent ease, but at the cost of a feeling of instability. The future is no longer a continuous line—it’s an uncertain flow where we constantly adjust our position, sometimes without an overall vision.
To this generalized plasticity is added a training effect. The faster contexts change, the more individuals are expected to adapt, to train themselves, to “recalibrate.” The promise of autonomy then merges with a demand for perpetual self-reinvention. We move constantly, but we no longer know very well toward what. This permanent fluidity produces a paradoxical form of immobility, a continuous displacement without a clear destination [4] [5].
3. The acceleration of the pace of life. This is the logical consequence of the first two. As we gain time on each gesture, we fill in more of them. Our days overflow. A revealing symptom: more and more people watch Netflix at accelerated speed (x1.5, even x2) to “save time.” Save time… on entertainment. Even our leisure has become a race against the clock. We listen to podcasts while walking, we answer emails during meetings, we check Instagram while eating. Multitasking is no longer an exceptional skill—it has become the minimum norm.
Every free moment becomes a slot to exploit, every pause an opportunity to “catch up” on an imaginary delay. We live in an economy of saturation where inaction is experienced as a moral fault. The famous to-do list never empties: the more boxes we check, the more appear. And at the end of the day, exhausted, we have the impression of having done everything except what’s essential.
This dynamic has something implacable about it. Rosa speaks of a structural necessity of acceleration: it’s no longer an individual option, but a condition of survival in a self-sustaining system. The company that doesn’t innovate fast enough disappears, the nation that doesn’t grow fast enough collapses. We turn in the wheel, like the hamsters Rosa describes, condemned to run endlessly simply to avoid being ejected from the movement. “No matter how fast you run this year,” he writes, “next year, you’ll have to go faster.”
French philosopher Paul Virilio, pioneer of dromology (the science of speed), had sensed it since the end of the 20th century: speed is not neutral, it is the first instrument of power [6]. He who controls the rhythm of others controls their world. In this perspective, slowing down is not just a question of well-being or personal ecology—it’s a political act. Resisting speed means refusing to let the cadence of one’s mind be dictated by others.
But this external acceleration is not without internal effect. By running so as not to fall, we’ve ended up forgetting why we were running. This continuous agitation shapes our perceptions, our relationship to the world and to ourselves. It pushes us to think fast, to judge without distance, to react rather than to understand. And that’s where the real dispossession begins: when the rhythm of the world ends up shaping that of our mind.
Because if everything accelerates around us, our brain hasn’t changed. It still functions according to two speeds, two logics, two systems. And it’s there, in this silent tension between instinct and reflection, between rapidity and accuracy, that our capacity to remain lucid is now at stake. To understand this, we must go down a notch: inside our heads, where the speed of the world collides with the millenary architecture of our thought.
Two-speed thinking
If acceleration is the tempo of the external world, our brain hasn’t changed. It continues to function according to two regimes of thought—one fast and intuitive, the other slow and deliberative. This distinction, formulated by psychologist and economics Nobel laureate Daniel Kahneman in Thinking, Fast and Slow [7], illuminates with striking precision the way contemporary technology diverts our attention and our judgments.
Imagine your mind as an organization with two directors who don’t always talk to each other. The first director, whom Kahneman calls System 1, reigns over automatisms. It acts without apparent effort, relies on emotions, impressions, habits. It’s the one that makes you recognize a familiar face, guess an intention in a look, or complete a sentence before it’s even finished. It excels in urgency, but it privileges the coherence of a story over the rigor of facts. It loves to simplify.
The second director, System 2, is slower, more methodical. It observes, calculates, verifies. It’s the one that kicks in when you need to solve a logic problem, make a reasoned choice, or question a first impression. It requires attention, energy, concentration. And because it tires quickly, it tends to give way as soon as possible to its impulsive counterpart.
Kahneman emphasizes that our brain is a cognitive miser: it saves its efforts whenever it can. That’s why we often let System 1 decide for us, even in situations that would deserve more reflection. Psychologist Shane Frederick demonstrated this through what has become a classic experiment: the Cognitive Reflection Test (CRT) [8]. He poses a simple question:
A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost?
Most people answer immediately: ten cents. Quick response, intuitive, wrong. If the ball cost ten cents, the bat would cost $1.10, and the total would be $1.20. The right answer, five cents, only appears to those who take the time to slow down, to set up an equation, to activate System 2.
This test is not a logic game, but a mirror of our relationship to thought. It shows how much we trust the first plausible idea, even when it’s erroneous. In a digital environment saturated with signals, notifications, and ready-to-consume content, this tendency intensifies. Each click, each like, each instant response reinforces the reign of System 1: that of reflex, of the instant, of immediate gratification.
Artificial intelligence, seemingly so rational, amplifies this drift. By delivering us well-crafted answers before we’ve even had time to fully formulate the question, it dispenses us from the effort of reasoning. Why linger on a problem if a machine solves it in our place? Why mobilize one’s critical thinking when the answer is fluid, coherent, elegant? Little by little, we stop activating System 2. The effort of thought becomes superfluous, almost archaic.
Yet a mechanism that is no longer exercised atrophies. What we believe we gain in efficiency, we lose in lucidity. By dint of thinking fast, we cease to think accurately. And this invisible fatigue of judgment, this laziness of discernment, is perhaps the real fragility of our time.
Because System 1 loves certainties, but it dreads doubt. And when a technology feeds it permanently with immediate and well-formulated answers, it’s our System 2 (the one of reflection, verification, judgment) that gradually falls asleep. This invisible fatigue of discernment, this cognitive laziness encouraged by the machine, is not without consequences. It leaves traces in our brain, traces that neuroscience is only beginning to measure.
Cognitive atrophy: when the brain shuts down
American writer Nicholas Carr was one of the first to sound the alarm. In an article with a deliberately provocative title, “Is Google Making Us Stupid?”, published in 2008 in The Atlantic, he described a personal experience that resonates today with troubling acuity:
“Immersing myself in a book or a lengthy article used to be easy. My mind would get caught up in the narrative or the turns of the argument, and I’d spend hours strolling through long stretches of prose. That’s rarely the case anymore. Now my concentration often starts to drift after two or three pages. I get fidgety, lose the thread, begin looking for something else to do. I feel as if I’m always dragging my wayward brain back to the text. The deep reading that used to come naturally has become a struggle.” [1]
This early confession gave birth to a book, The Shallows (translated as Does the Internet Make Us Stupid?), where Carr demonstrates, with neuroscience to back him up, that our brain is plastic and reconfigures itself according to the tools we use. By favoring skimming, multitasking, and constant interruption, the Internet unlearns us from deep, linear, and concentrated reading. This phenomenon is aggravated by what researchers called as early as 2011 the “Google effect”: when we know that information is easily accessible online, our brain makes less effort to memorize it [9]. We externalize our memory, entrusting it to distant servers, and the neural circuits dedicated to memory consolidation, for lack of being solicited, weaken.
If the Internet was already reshaping our brains, the advent of generative artificial intelligence seems to be acting as a powerful accelerator of this transformation. A study conducted by researchers at the Massachusetts Institute of Technology (MIT), published in 2025, attempted to measure its impact. Its findings have stirred up considerable controversy in the scientific community, as they appear deeply unsettling.
The experiment, led by neuroscientist Nataliya Kosmyna [24], focused on the ability to write an essay—a task that involves memory, but not exclusively. Around fifty students from the Boston area were asked to write a text in twenty minutes, either without assistance, with Google, or with ChatGPT. During the exercise, their brain activity was recorded via electroencephalography, and participants were then asked questions about their own text and had to recall certain passages from it.
The results were striking: the vast majority of those who had used ChatGPT couldn’t remember any passage from what they’d produced, whereas the two other groups could do so easily. Analysis of the brain signals revealed significantly lower activity among AI users, particularly in areas associated with episodic memory. In other words, they had formed no memory whatsoever related to the writing task.
“It’s not really surprising,” comments Nataliya Kosmyna, “because how can you remember a text if you didn’t write it?”
What the study demonstrates, then, isn’t a failure of memory itself, but rather a decline in cognitive engagement: participants were outsourcing not just the task, but also the mental effort that goes with it. And it’s precisely this effort—the brain’s work to organize, formulate, and reformulate—that creates genuine memory. Without it, the memory doesn’t take hold; it vanishes before it can even exist..
The data collected suggests that ChatGPT users certainly drafted their texts 60% faster, but at the cost of a 32% drop in what researchers call “relevant cognitive load.” Let’s stop for a moment on this term. It designates the intellectual effort necessary to transform raw information into structured and personal knowledge. It’s not simply reading or hearing something—it’s the mental work that follows: reformulating the idea in one’s own words, connecting it to what one already knows, identifying what one hasn’t understood, filling in the gaps. It’s this effort, often invisible, that builds our real understanding. Without it, information slides over us without transforming us.
With ChatGPT, this work disappears. The machine delivers you a text already structured, already formulated, already “understood.” You just have to read it, approve it, copy it. The brain, deprived of this translation effort, memorizes nothing, understands nothing deeply. It’s like eating a meal already chewed: it fills the stomach, but it doesn’t nourish.
The electroencephalogram (EEG) would have revealed a significant decrease in brain connectivity among AI users. One figure is particularly striking: 83% of them would have been unable to remember a passage they had just had written by the machine [3]. While these results deserve to be confirmed by other research, they nevertheless point to an essential question: what happens in our brain when we delegate the effort of thinking?
What this study, despite its limitations, allows us to glimpse is the concept of cumulative cognitive debt. By systematically delegating mental efforts (structuring a thought, finding the right words, constructing an argument) to AI, we risk contracting a debt to our own brain. The prefrontal cortex, seat of higher cognitive functions, could be increasingly less solicited. Like a muscle that is no longer used, it atrophies. And this potential atrophy would have lasting effects that go far beyond the immediate task, affecting our ability to think critically and creatively in all aspects of our life.
This cognitive laziness encouraged by AI leads to another perverse phenomenon: the uniformization of thought. A British study showed that while ChatGPT could indeed improve the quality of individual texts, the overall creativity of a group of authors using the tool decreased [10]. Texts became more homogeneous, less original. In parallel, a vast study conducted by Microsoft on knowledge workers revealed a substantial negative correlation between the frequency of AI tool usage and critical thinking scores [11]. Even more troubling, the study shows that this “cognitive offloading” amplifies when the user’s trust in the model exceeds trust in their own skills. We then stop making the effort to doubt, to verify, to criticize, contenting ourselves with the plausible and well-formulated answer from the machine.
This erosion of critical thinking is accompanied by non-negligible psychological risks. The very nature of generative AIs (which express themselves like humans, adapt to our behaviors, and seem to have an answer for everything) has everything to make us dependent. This dependence can lead to social isolation, reflective disengagement (“Why learn or think for myself if a machine can do it?”), and even a deep feeling of humiliation in the face of the crushing efficiency of these tools [2].
The picture is dark, but it’s not fatal. The acceleration of the world pushes us toward fast and superficial thinking. The tools we’ve created seem precisely designed to atrophy the circuits of slow and deep thought. Yet this atrophy is not inevitable. It’s not a disease, but a neglect. And any neglect can be corrected.
On the condition of cultivating two virtues that the modern world despises: doubt and lucidity. Not the paralyzing doubt that leads to cynicism, but the methodical doubt that opens thought. Not the arrogant lucidity that thinks it knows everything, but the humble lucidity that knows how to look without being blinded. These two qualities, sober but powerful, form the foundation of all intellectual resistance. They are the primary tools to regain control over our own mind.
Doubt and lucidity, or how to cultivate critical thinking
After measuring the extent of the problem (this acceleration that shapes the world, this two-speed brain that lets itself be trapped, this atrophy that threatens us), it’s time to explore the paths of resistance. How, concretely, can we cultivate daily this vigilance that prevents automatism from engulfing us?
Thinking slowly means first accepting to pause. In the tumult of notifications, instant debates, and certainties thrown online as self-evident truths, there is this almost subversive gesture: to doubt. Not to doubt out of weariness or cynicism, but out of rigor. Doubt is not a crack in thought—it’s a skylight opened onto reality. It forces us to breathe before responding, to question before believing. In a world where everything accelerates, doubting becomes a form of asceticism, a gentle resistance to haste. It’s a return to slowness, a way of letting truth come at its own pace, without rushing it. But for this doubt to become fertile, one must still learn to think methodically, to distinguish rigor from distrust: that’s where critical thinking and critical spirit meet.
Critical thinking and critical spirit
But what exactly are we talking about when we evoke critical spirit? The expression may circulate everywhere, but it often remains vague, reduced to a posture of general distrust or a simple reflex of contradiction. In reality, critical spirit is not a posture of rejection—it’s a rigorous way of examining what to believe or do based on reasons, in a reasonable and thoughtful manner, as Robert H. Ennis formulates it [15]. Matthew Lipman specifies this dynamic: it’s a judgment based on criteria, capable of self-correction and attentive to context [16]. John McPeck speaks of a skill coupled with a propensity to practice reflective skepticism, while Harvey Siegel insists on its culmination in action, thinking and acting appropriately because founded on reasons [17] [18].
But we must distinguish critical thinking and critical spirit. The first refers to intellectual skills: analyzing an argument, evaluating a source, formulating objections. The second designates the dispositions that make these skills truly operative, the will to activate them, the patience to bear their slowness, the honesty to draw their consequences. This is the point emphasized by Jacques Boisvert: mastering the evaluation of reasons is not enough if we don’t cultivate the attitudes that carry this exercise daily [19].
In other words, critical spirit is an intellectual habitus—that is, a stable disposition of the mind, acquired through repeated exercise of judgment. Pierre Bourdieu used this term to designate those ways of thinking, perceiving, and acting that become so deeply rooted that they become natural. The habitus here is therefore not an automatic reflex, but a form of cultivated intuition, the fruit of a slow maturation where vigilance becomes second nature.
Let’s take a concrete example. You come across an article massively shared on social networks. A shocking figure, an alarmist headline. At the beginning of your critical apprenticeship, you had to force yourself to verify: “What’s the source? Who published this study? Are the figures contextualized?” It was a conscious, tiring effort. But by repeating this gesture, something changes. Doubt no longer arises after the fact—it’s already there, discreet but present, as soon as you read the title. You no longer tell yourself “I need to verify”—you simply verify, like a musician improvises without thinking about the notes. Critical spirit has become an intelligent reflex, a learned spontaneity. It’s no longer an effort against laziness—it’s a way of being.
Critical spirit combines know-how (analyze, compare, verify) and know-how-to-be (vigilance, openness, constancy). In the age of instant answers, it demands a small disobedience to the ambient rhythm: accepting the time of doubt, preferring tested coherence to seductive ease. It’s this alliance of capacities and dispositions that will allow us, in what follows, to illuminate the limits of AIs and the necessity of active lucidity.
Why doubt is necessary
Doubt is not the enemy of knowledge—it’s its condition. It’s what prevents thought from freezing, maintains vigilance, and protects the mind against the ease of evidence. René Descartes already made it the starting point of all true certainty: one must suspend judgment to better distinguish the true from the plausible. But in a world saturated with instant answers, doubt is no longer a method, it becomes an act of resistance.
Resist against what? Against speed, first. Against the temptation of everything, right now. To doubt is to accept slowing down when everything pushes you to react. It’s refusing to be sucked in by the cadence of flows, by this succession of affirmations that no longer leaves room for silence. To also resist conformity, the kind that makes consensus a substitute for truth. Doubt is the space of counter-tempo, the right to examine before adhering, the courage to go against the grain of dominant opinion.
Artificial intelligence, through its fluidity and self-assurance, tends to erase this effort. It produces such well-crafted sentences, such logical sequences, that they disarm us. Everything appears coherent, as if truth no longer depended on the work of discernment, but on surface elegance. The danger doesn’t lie in error (machines make mistakes, but correct themselves)—it lies in our loss of the verification reflex. Doubt, this inner gesture that consisted of examining, comparing, cross-checking, fades behind the immediate satisfaction of a well-formulated answer.
To doubt, today, is to accept becoming slow again. It’s questioning the origin of an idea, the solidity of a proof, the coherence of a reasoning. It’s also repaying, act by act, the cognitive debt we’ve accumulated. It’s refusing the temptation of the mental shortcut to rediscover the flavor of the intellectual path. In a sense, doubt is a form of politeness toward truth: it doesn’t rush it, it lets it come. And this patience, in a world of automatisms, has become a rare virtue, perhaps even the first form of freedom.
What does it mean to be lucid in the age of AI?
But what exactly are we talking about when we invoke lucidity? The word might seem austere, almost clinical. Yet it designates something both simple and fundamental: a certain quality of gaze. To be lucid is not to sink into unhealthy distrust, nor to barricade oneself against technology. It’s learning to see without being dazzled. Science journalist Florian Gouthière offers a sober and accurate definition: “an acquired capacity allowing one to evaluate different aspects of information before forming an opinion about it—essentially, regarding the level of confidence one can accord it” [20]. In other words, lucidity doesn’t reject—it weighs. It doesn’t condemn—it questions.
In the case of artificial intelligences, this lucidity begins with a form of salutary disenchantment. We must accept seeing these machines for what they really are: not infallible oracles, but linguistic prediction systems of formidable efficiency. They assemble words according to probabilities—they don’t think. Recognizing their utility without attributing virtues they don’t have is the first gesture of lucidity. And this understanding shouldn’t remain the prerogative of engineers. It falls under what we now call digital literacy: this aptitude to handle technological information with discernment and critical distance [21].
But being lucid also means turning one’s gaze inward. It’s accepting to recognize our own flaws, this propensity our System 1 has to prefer apparent coherence to laborious truth. We like to believe we’re rational, masters of our judgments. Lucidity reminds us of a humbler reality: we are all, without exception, sensitive to the attraction of ease. Machines speak so well, so quickly, with such assurance, that they disarm our vigilance. We must therefore cultivate this resistance, not as a natural gift, but as a deliberate, patient, repeated effort. Digital scientist Aurélie Jean sums it up thus: developing a scientific culture helps build one’s critical spirit, by privileging questioning, structuring reflection, constructive criticism [22].
Finally, lucidity possesses an even more political, more essential dimension. It invites us to never forget that behind the apparent neutrality of algorithms lie human choices, selected data, defined objectives, encoded values. Each generated response carries with it economic interests, ideological orientations, invisible power relations. To be lucid is to systematically ask this question: who profits from this answer? Who does it really serve? We find here Hannah Arendt’s deep intuition, for whom thinking was first and foremost a civic responsibility. Refusing to swallow without thinking what is served to us (whether by a totalitarian regime or by a chatbot interface) is refusing to be reduced to the state of a docile executor. It’s affirming that one remains a thinking subject, capable of autonomous judgment.
The concrete gestures of lucidity
This lucidity we speak of doesn’t float in abstraction. It translates into concrete gestures, daily practices that, put end to end, constitute a true hygiene of the mind. These gestures fall under what media education researchers call critical digital literacy, this competence that has become vital “for the security, health and well-being of individuals and communities” in the age of digital disinformation [23].
Verify before relaying. When an AI generates information, a fact, a statistic, the lucid reflex consists of verifying it before sharing or using it. Cross-check with reliable sources, go back to the primary source, consult human experts. This simple gesture, which takes just a few minutes, forms a rampart against the spread of disinformation and against our own credulity. It’s one of the fundamental capacities identified by Ennis: “evaluating the credibility of a source” [15].
Systematically confront an automated response with a human reading. When you solicit an AI to write a text, synthesize a document, or analyze a situation, never be content with the first answer. Reread carefully, question the logic, look for blind spots. Better yet, confront this answer with that of a colleague, a friend, a book. Human thought, with its imperfections and slownesses, brings a depth and nuance that the machine cannot simulate. This practice corresponds to what Ennis calls “analyzing arguments” and “formulating clarification or challenge questions” [15].
Practice voluntary “latency time.” When you receive an instant response from an AI, impose a delay on yourself before using it. Let it rest, come back to it a few hours or a few days later with a fresh eye. This latency time allows your System 2 to regain control, to spot inconsistencies or approximations that initial enthusiasm had masked. It’s a deferred repayment of cognitive debt: what you gained in speed, you recover in depth.
Document your sources. Always note where information comes from, whether it’s generated by an AI or found elsewhere. This traceability allows you to go back, to verify, to correct. It’s also an exercise in humility: it reminds us that our thought is never entirely autonomous, that it always relies on a network of sources that must be honored and examined.
Cultivate diversity of sources. Never rely on a single AI model, a single search engine, a single perspective. Multiply viewpoints, consult media of different sensibilities, read authors who disturb you. This diversity is the best antidote to the uniformization of thought and to confinement in filter bubbles. It joins Lipman’s definition of critical thinking that is “permeable to context” [16].
These gestures may seem modest, almost derisory in the face of the power of machines. But it’s precisely in this modesty that their strength resides. They require neither advanced technical skills nor specialized training. They simply demand vigilance, patience, and a commitment never to entirely cede the helm of our thought. Lucidity, at bottom, is that: keeping your hands on the wheel, even when the car offers to drive all by itself. It’s what McPeck called “the skill and propensity to engage in an activity with reflective skepticism” [17], not systematic rejection, but constant vigilance, a disposition to question that becomes second nature.
The philosophy of slowness: relearning how to think
These concrete gestures of lucidity already sketch a form of resistance. But they’re not sufficient by themselves. Because beyond techniques and protocols, it’s a philosophy of life that must be reinvented. A way of inhabiting time differently, of reappropriating one’s own inner duration. Fortunately, this counter-culture of slowness doesn’t need to be invented: it already exists, carried by thinkers who, well before the age of AI, knew how to identify the poison of speed.
Hannah Arendt and the “stop-and-think”
At the heart of this intellectual resistance stands the figure of Hannah Arendt. More than anyone, she probed the dangers of the absence of thought. By analyzing totalitarianism and observing the trial of Adolf Eichmann, she forged this terrifying idea of the “banality of evil”: monstrous acts committed not by monsters, but by zealous functionaries, ordinary men incapable of thinking for themselves, incapable of stopping to question the nature of their actions. For Arendt, there exists a “strange interdependence between the absence of thought and evil” [12].
She warns us against a confusion that nevertheless seems natural to us: that between thinking and knowing. Knowledge seeks certainties, facts, definitive answers. Thought is something else entirely, an incessant inner dialogue, a questioning that exceeds the simple accumulation of knowledge. To think is to “stop and think” (stop-and-think), to interrupt the flow of activities to seek meaning. It’s this capacity to pose “all the unanswered questions on which every civilization is founded” that constitutes, for Arendt, our greatest political responsibility.
Margarethe von Trotta’s film about Hannah Arendt magnificently illustrates this requirement of slowness. It shows us long sequences where the philosopher, cigarette in hand, seems to do nothing. She thinks. The film stages the tension between the demands of The New Yorker, which awaits her article on the trial with the impatience of the journalistic world, and the long, incompressible time of philosophical reflection. Arendt’s work, the film tells us, “is the kind of work that cannot be hurried” [12]. In our world of tweets and instant reactions, Arendt reminds us that haste is the enemy of thought. “Haste tweets without thinking, and without considering the consequences,” as Michelle Boulous Walker, a contemporary heir to Arendt’s thought, aptly formulates [12].
Michelle Boulous Walker and Slow Philosophy
It’s precisely Michelle Boulous Walker who theorized the concept of “Slow Philosophy.” For her, philosophy is, by essence, “the art of reading slowly” [12]. She opposes the “corporatization of thought,” this logic of efficiency and productivity that has invaded even the academic world and reduces thought to simple information extraction. Slow philosophy, on the contrary, is an “attention to thought,” a “rare and intense meditation that transforms us from one state to another” [12].
Walker invites us to question the negative connotations we’ve ended up associating with the word “slow.” Since the industrial revolution, slowness has become synonymous with stupidity, inefficiency, boredom. Slow Philosophy contests this hierarchy and operates a vital distinction: speed, which can sometimes be necessary and desirable, should not be confused with haste, which is always harmful. Haste is a negligent and superficial relationship to the world; it fails to grasp complexity and undermines our capacity for judgment. Slowness, on the other hand, is the soil on which deep thought can germinate.
Carl Honoré and the Slow Movement
This idea was popularized well beyond philosophical circles by journalist Carl Honoré, author of the worldwide bestseller In Praise of Slowness [13]. Honoré became the spokesperson for the Slow Movement, a counter-culture that touches all aspects of life, from food (Slow Food) to education, including work and leisure. The idea, he insists, is not to “do everything at a snail’s pace.” Rather, it’s about finding the tempo giusto, the right tempo. It’s a call for a better balance between speed and slowness, to reappropriate our time to improve our quality of life. Slow Living is not a simple time management technique—it’s a philosophy of life: anchoring oneself in the present moment, feeling one’s environment with more sensitivity, living more consciously, and recreating authentic bonds, with others and with oneself.
Frédéric Gros and the philosophy of walking
There exists a particularly pure incarnation of this philosophy of slowness: walking. Not athletic walking, measured in kilometers and calories burned, but walking as an experience of thought. This is what philosopher Frédéric Gros explores in his essay A Philosophy of Walking [14]. “Walking is not a sport,” he warns. It’s neither a performance nor an exploit. Walking is something else, a parenthesis in the urgency of the world, a space where body and mind rediscover their original tempo.
To walk is to accept a form of radical simplicity that tears us away from ambient utilitarianism. You produce nothing by walking, you consume nothing. You’re simply there, present to the world that slowly scrolls around you. This presence offers the body what Gros beautifully calls “a moment of parole,” a precious interval where thought can wander freely, unfold without constraint, deepen to the regular rhythm of steps. Between the frenetic speed that makes us dizzy and the immobility that stiffens us, walking opens a third way.
From Arendt to Gros, passing through Walker and Honoré, these different approaches converge toward the same intuition: to truly think, one must first know how to stop. Slowing down is not an admission of weakness—it’s an act of strength, one that allows us to cultivate the most precious tool in the face of the complexity of the world and the opacity of machines: our critical mind.
Slowness as resistance and as luxury
We’ve reached the end of the journey. We’ve crossed speed and its mirages, probed the corners of a mind saturated with algorithms, encountered the voices of Arendt, Kahneman, and Rosa, like beacons in the fog. And at the end of this crossing, a feeling remains: thought cannot be rushed without losing itself. It’s a fragile beat, an interval between two flows, a breath that claims time to exist.
We believed that rapidity would make us more efficient, more lucid, more powerful. But efficiency hasn’t kept its promises: it has gnawed at depth, flattened doubt, dulled memory. We’ve lived on intellectual credit, and the time has come to repay what we owe to our own mind. What the machine accomplishes brilliantly, it doesn’t understand; and what we once understood, we sometimes cease to inhabit. Artificial intelligence is not our enemy—it’s our reflection. It calculates what we entrust to it, amplifies what we flee from: our fear of waiting, our vertigo of silence, our fatigue of thinking.
To slow down, therefore, is no longer simply mental hygiene, but a gentle insurrection. It’s a refusal of reflex, a gesture of reconquest. It’s sitting at the edge of the tumult and making contact again with the slowness of the world. It’s not turning one’s back on technology—it’s opposing it with a human rhythm, that of discernment, of groping, of nuance. To think slowly is to restore the right distance between question and answer, between evidence and truth. It’s remembering that lucidity is not a lightning bolt, but a vigil.
And perhaps, to understand what’s at stake, we must return to that original scene we evoked at the beginning: Bowman facing HAL. The man, methodical, disconnects the circuits of the machine gone mad with coherence. It’s not vengeance—it’s an act of love: returning silence to consciousness. Each wire he removes seems like a thought regained, each light that goes out an illusion that dissolves. What he’s disconnecting, in truth, is not the machine: it’s within himself the temptation of speed, the pride of control, the fear of emptiness.
Perhaps by also disconnecting our automatisms and our urgencies, we could relight that discreet breath still called thought. Not to turn off the machine, but to relearn how to breathe with it, in another rhythm, in another listening.
Because true humanity is not measured by the power of our tools, but by the slowness of our awakenings. And in a world that confuses movement with meaning, slowing down is no longer a weakness, but a courage: that of remaining present to oneself, and alive among the machines.
References
For meticulous minds, lovers of numbers and sleepless nights verifying sources, here are the links that nourished this article. They remind us of one simple thing: information still exists, provided one takes the time to read it, compare it, and understand it. But in the near future, this simple gesture may become a luxury, because as texts generated entirely by AI multiply, the real risk is no longer disinformation, but the dilution of reality in an ocean of merely plausible content.
[1] Carr, Nicholas. “Is Google Making Us Stupid?”. The Atlantic, July/August 2008, https://www.theatlantic.com/magazine/archive/2008/07/is-google-making-us-stupid/306868/.
[2] Roxin, Ioan. “Generative AI: the risk of cognitive atrophy”. Polytechnique Insights, July 3, 2025, https://www.polytechnique-insights.com/tribunes/neurosciences/ia-generative-le-risque-de-latrophie-cognitive/.
[3] Massachusetts Institute of Technology (MIT) study on the use of ChatGPT and cognitive load, cited by Ioan Roxin in Polytechnique Insights.
[4] Rosa, Hartmut. Acceleration: A New Critical Theory of Modernity. Columbia University Press, 2013.
[5] “Hartmut Rosa’s Social Acceleration Theory: A New Critical Theory of Modernity”. Medium, August 5, 2023, https://medium.com/@s-blog/hartmut-rosas-social-acceleration-theory-a-new-theory-of-modernity-ac8ef9fef799.
[6] Virilio, Paul. Speed and Politics. Semiotext(e), 1977.
[7] Kahneman, Daniel. Thinking, Fast and Slow. Farrar, Straus and Giroux, 2011.
[8] Frederick, Shane. “Cognitive Reflection and Decision Making”. Journal of Economic Perspectives, vol. 19, no. 4, 2005, pp. 25-42, https://www.aeaweb.org/articles?id=10.1257/089533005775196732.
[9] Sparrow, B., Liu, J., & Wegner, D. M. “Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips”. Science, vol. 333, no. 6043, 2011, pp. 776-778.
[10] British study on group creativity with ChatGPT, cited by Ioan Roxin in Polytechnique Insights.
[11] Microsoft study on critical thinking and AI, cited by Ioan Roxin in Polytechnique Insights.
[12] Walker, Michelle Boulous. “Why slow philosophy is the antidote to fast politics”. ABC Religion & Ethics, June 9, 2020, https://www.abc.net.au/religion/michelle-boulous-walker-slow-philosophy-in-a-time-of-fast-polit/12336408.
[13] Honoré, Carl. In Praise of Slowness. HarperOne, 2004.
[14] Gros, Frédéric. A Philosophy of Walking. Verso, 2014.
[15] Ennis, Robert H. “Critical Thinking: A Streamlined Conception”. Teaching Philosophy, vol. 14, no. 1, 1991, pp. 5-24.
[16] Lipman, Matthew. Thinking in Education. Cambridge University Press, 1991.
[17] McPeck, John E. Critical Thinking and Education. St. Martin’s Press, 1981.
[18] Siegel, Harvey. Educating Reason: Rationality, Critical Thinking and Education. Routledge, 1988.
[19] Boisvert, Jacques. La formation de la pensée critique : Théorie et pratique. De Boeck, 1999.
[20] Gouthière, Florian. Santé, science, doit-on tout gober ?. Belin, 2017.
[21] “Digital Literacy”. Centre de documentation sur l’éducation des adultes et la condition féminine, https://cdeacf.ca/dossier/litteratie-numerique.
[22] Jean, Aurélie. “Developing critical thinking through scientific culture”. Cited in various publications on digital literacy.
[23] “Digital Media Literacy and Critical Thinking Online”. U.S. Department of Homeland Security, https://www.dhs.gov/sites/default/files/publications/digital_media_literacy_1.pdf.
[24] Nataliya Kosmyna *Research Scientist at MIT Media Lab https://www.braini.io/bio