Protecting Ourselves from Ourselves

AI is taking over critical thinking tasks

In the waning years of the Roman Republic, Cicero stood in the Senate, his voice a resounding appeal against the erosion of reasoned debate. He warned that a society unwilling to question the claims of its leaders would soon become captive to them. History vindicated his fears. The Republic crumbled, not through external conquest, but from within, as rhetoric replaced truth, spectacle supplanted substance, and the citizens, lulled by the strong voice of the 'anointed' leader and his propaganda, ceased to interrogate power. It is a lesson as old as civilization, the fate of a society is tethered to its ability to think critically.

The ability to question, to think critically, is civilization’s immune system. To think critically is not to merely doubt, but to wield doubt constructively, to interrogate ideas, not dismiss them, to challenge assumptions, not accept them unexamined. It is an art form that demands intellectual rigor, the disciplined application of reason and evidence in evaluating claims and arguments, and, above all, the courage to remain unsatisfied with easy answers.

Yet today, in an era saturated with information, genuine understanding has never been more elusive. A 2021 Pew Research study found that while 93% of Americans report consuming news daily, only 26% feel confident in distinguishing fact from opinion. We mistake the abundance of data for a wealth of knowledge, and the relentless churn of news for informed discourse. But information without scrutiny is a journey without direction, pushing us not toward clarity, but toward confusion. The modern public sphere, shaped by algorithms that prioritize reaction over reflection, has become an arena where engagement is rewarded over enlightenment. If we are not vigilant, we risk surrendering to a new form of intellectual serfdom, not at the hands of tyrants, but of our own inattentiveness.

Historically, societies that embraced rigorous questioning flourished. The Renaissance did not emerge from passive acceptance, but from a defiant challenge to dogma. Galileo’s insistence that empirical observation must triumph over ecclesiastical decree was not merely a scientific breakthrough, it was a declaration that no idea, no matter how sacred, should be immune to scrutiny. Likewise, the Enlightenment was built on a foundation of doubt, a collective refusal to accept inherited truths without examination. It is no coincidence that these periods of intellectual vitality coincided with extraordinary advancements in human progress.

But critical thought is not merely an academic virtue, it is the foundation of ethical governance, just economies, and resilient institutions. In business, the difference between success and ruin often hinges on the ability to distinguish robust data from seductive but misleading statistics.

The strength of a democracy is not measured by the volume of voices, but by the quality of debate, by the extent to which policies are scrutinized, assumptions questioned, and evidence demanded. In technology, where algorithms increasingly shape human behavior, understanding the biases encoded within them is not just advisable, it is imperative.

This is not a call for skepticism alone. Unchecked cynicism leads to paralysis, the refusal to believe in anything can be as dangerous as blind faith. The goal is not to reject all claims, but to assess them, to refine our ability to separate truth from falsehood, logic from manipulation. To cultivate this skill requires more than intelligence, it demands discipline. One must read deeply, think slowly, and resist the seduction of simplistic narratives. It requires an embrace of complexity, an appreciation for ambiguity, and an understanding that certainty is often the enemy, I much prefer probabilities.

There are, of course, counterarguments. Some claim that in moments of crisis, prolonged questioning impedes decisive action. The need for swift, resolute decision-making, whether in war, emergency response, or financial collapse, is undeniable. However, history suggests that decisions untempered by critical thought can lead to catastrophe. The challenge, then, is not to abandon critical inquiry but to refine it, to develop a capacity for rapid yet rigorous evaluation, ensuring that urgency does not excuse unthinking acceptance.

Perhaps the greatest paradox of critical thought is this, the more rigorously we question, the less certain we become. But this is not a weakness, it is a sign of intellectual maturity. A mind unshaken by doubt is a mind unwilling to grow. Certainty breeds stagnation, but curiosity, restless, insatiable curiosity, propels us forward. And therein lies our greatest hope.

The world does not need more passive consumers of information, it needs architects of understanding. It needs those willing to sift through the noise, challenge prevailing assumptions, and, when necessary, stand against the tide of collective delusion. It is not an easy path. But then, nothing of value ever is.

As we stand at yet another historical precipice, faced with forces eager to manipulate and mislead, we must remember this, the future will not be shaped by those who accept, but by those who question. In that questioning, we reclaim not only our agency, but the very essence of what it means to be free.

This brings me to AI and critical thinking

A recent study, appropriately powered by AI and authored by researchers at leading AI Lab, Anthropic, offers a sweeping empirical analysis of how AI systems like Claude are being wielded in the workforce. The results are both unsurprising and deeply unsettling: AI is now an intellectual co-pilot, heavily concentrated in tasks demanding software development, technical writing, and data analysis.

The study finds that “cognitive skills like Reading Comprehension, Writing, and Critical Thinking show high presence” in AI-assisted tasks. This should alarm us. These are not mere clerical functions; they are the bedrock of human inquiry. If these skills are now routinely outsourced, how long before they atrophy like an unused muscle? Already, I see the telltale signs, the glazed eyes of students accustomed to instant answers, the hesitation of professionals who once prided themselves on their analytical prowess, now instinctively reaching for an AI-generated shortcut.

Of course, the optimists argue that AI is merely another tool, no different from a calculator or a word processor. But this is a facile comparison. The calculator does not think for you, it merely expedites computation. AI, however, thinks for you in a manner eerily close to cognition. The efficiency gains are undeniable, business professionals craft polished emails in seconds, software engineers debug code with a single prompt, consultants draft reports with the ease of conjuring a genie. But at what cost? Efficiency, as historian Lewis Mumford reminds us, is not synonymous with wisdom.

The study notes that while AI is still minimally used in physically intensive occupations, construction, anesthesiology, maintenance, its dominion over cognitive tasks grows with each passing month. In the past, technological advancements displaced physical work while leaving cognitive work intact. AI, for the first time, threatens to invert this equation. It is the thinker, not the doer, who now stands on the precipice of irrelevance.

It is tempting to see augmentation as benign, even beneficial. After all, what is wrong with having an always-available brainstorming partner? But augmentation is a mirage if it fosters dependence rather than mastery. The philosopher Ivan Illich warned of “shadow work”, the unseen ways in which technology transforms human behavior. AI augmentation may not displace thought immediately, but it subtly alters our intellectual habits, making us comfortable with surface-level answers, uneasy with complexity, impatient with the slow burn of difficult problems.

AI is not a mere disseminator of information. It is an expert, an ever-present interlocutor, and in its ubiquity, it threatens to invert literacy, from an active engagement with knowledge to a passive consumption of prepackaged insights.

The most intellectually vibrant sectors, education, research, technical analysis, are precisely where AI’s footprint is heaviest. It is in these domains that we must be most vigilant, lest our brightest minds become conduits for algorithmic reasoning rather than sites of original thought.

What, then, is the solution? First, we must resist the seduction of automation, particularly in the domain of critical thinking. The classroom, the laboratory, the boardroom, these must remain spaces where struggle is not an inconvenience but a virtue, of course we will use AI in these areas but they must be consciously used. Second, we must treat AI as a tool, not an oracle. The ease with which AI generates text, summarizes arguments, and refines logic should not lull us into intellectual complacency. Every AI-generated output must be interrogated, challenged, rewritten by a human hand.

Most importantly, we must cultivate intellectual resilience. In an age of instant answers, we must reclaim the art of difficult questions. The historian Richard Hofstadter once lamented the rise of anti-intellectualism in American life, today, the greater threat may not be hostility to intellect, but its passive surrender to machines. The measure of human intelligence has never been the ability to retrieve information, it has always been the capacity to question, synthesize, and create. If we outsource that, we outsource what makes us human.

What happens when augmentation becomes atrophy? When the crutch of AI becomes the replacement of intellectual rigor?

Stay curious and keep asking questions

Colin

Image at the top. I posted this essay into Gemini and asked it to create an image which represents the essay - I will say no more!

Note on Cicero: The evidence for Cicero's warnings about the decline of the Roman Republic comes from historical records, primarily his own writings, such as The Philippics and De Re Publica, as well as contemporary accounts by Sallust and later analyses by historians like Edward Gibbon. Cicero argued that the erosion of reasoned debate, the manipulation of public sentiment, and the increasing power of authoritarian figures, most notably Julius Caesar, were signs of Rome’s democratic decay. The eventual collapse of the Republic into autocratic rule under Augustus vindicated Cicero’s concerns, demonstrating how unchecked rhetoric and weakened institutional checks led to the end of self-governance.