The Risk of Frictionless Thinking

“The tools that amplify cognition can also hollow out the very capacities required to use them wisely.”

This quote from tech philosopher Tom Chatfield may be one of the defining paradoxes of the AI era.

Large language models give us extraordinary cognitive power. They help us write, summarize, reason, translate, plan. They save time. They reduce friction. They make us faster, more fluent, more capable.

And yet, the more we rely on them without reflection, the more we risk eroding the very skills we need to use them well: critical thinking, judgment, autonomy, and reflective reasoning.

AI makes us stronger. But only if we remain capable of thinking for ourselves.

Philosopher Hannah Arendt warned that the greatest danger to human judgment is not ignorance, but thoughtlessness, the failure to stop and think. For Arendt, thinking is not about producing answers, it is about maintaining a space of reflection where responsibility can exist. She sees thinking is an “internal dialogue.” AI doesn’t just provide answers; it potentially interrupts the internal monologue required for moral life. So, if the AI provides the “voice,” our own inner voice goes quiet. When that space collapses, individuals may still act efficiently, but no longer autonomously.

AI accelerates the risk of losing our autonomy of thinking. By delivering fluent outputs instantly, it tempts us to skip the pause where judgment is formed.

If answers arrive instantly, fully formed, and confidently phrased, what happens to the slow work of questioning, doubting, reasoning, and deciding?

Socrates never taught by giving answers. He taught by asking questions, relentlessly. His method assumed something essential: that thinking is not a product to be consumed, but an activity to be practiced. Thinking happens through dialogue, tension, uncertainty, and effort.

Machines can now generate answers. They cannot replace that effort.

“Learning comes with a cost. You have to mobilize cognitive, emotional, and social resources — there is no learning without effort.” Professor Grégoire Borst

Neuroscientist Gregoire Borst offers a crucial reminder: learning is inherently costly. It requires cognitive, emotional, and social effort, and there is no shortcut around that cost. When tools remove friction entirely, they don’t just save time; they risk removing the very conditions under which learning occurs. As Borst puts it, without a solid base of knowledge, critical thinking becomes impossible. And in a world where AI produces fluent answers that blur facts and beliefs, the effort to question, verify, and judge becomes more essential than ever.

The danger is that AI will make us less reflective. If we delegate thinking to machines, trusting outputs without interrogation, accepting fluency as truth, speed as wisdom, we risk losing something deeply human: the capacity to judge for ourselves, to take responsibility for our decisions, to remain authors of our own thoughts

Consumer LLMs are designed for helpfulness. They anticipate our needs, smooth the path to an answer, and remove friction wherever possible. But in education, friction is not a flaw: it is the point. Learning happens in what researchers call productive struggle: the effortful process of grappling with uncertainty, testing ideas, and revising one’s thinking. When obstacles disappear too quickly, understanding rarely has time to form.

This is why philosophy must move to the center of education.

Philosophy, in its essence, is not abstract or elitist. It is one of the most practical disciplines we have. It trains students to question assumptions, to evaluate arguments, to tolerate uncertainty, to think independently, precisely the skills required in an AI-saturated world.

France offers an example, with philosophy being mandatory in the final year of high school. Yet today students do not wait until then to use AI. If machines now produce effortless answers, education must teach, much earlier, the effort of thinking: how to question outputs, resist intellectual laziness, reason in everyday situations, doubt about an authority and exercise moral courage when judgment cannot be delegated. This is not enrichment; it is intellectual self-defense.

The issue is not only whether philosophy is taught, but how and when it is introduced.

That is the paradox of AI: an unprecedented increase in cognitive power without a corresponding increase in wisdom. Philosophy is not a commentary on this paradox; it is an education to judgment, teaching us to meet power with discernment, efficiency with wisdom and technological capability with meaning.

Svenia Busson is the co-founder of Education Futures

Adélaïde de Lastic is a French philosopher and consultant specializing in corporate ethics and values. She holds a doctorate in philosophy from the École des Hautes Études en Sciences Sociales (EHESS) and is an associate researcher at the Institut Jean Nicod (CNRS, EHESS, ENS).

Thanks for reading, to go further we recommend our podcast episode with philosopher Tom Chatfield “How to think critically in the age of AI” available on Spotify, Apple Podcast and Youtube

as well as the episode with neuroscientist Gregoire Borst “What is AI doing to our kids’ brains?” available on Spotify, Apple Podcast and Youtube.