In the modern age of information, where nearly every question can be typed into a search engine or asked of an AI chatbot, we would expect truth to be more accessible than ever. The entire repository of human knowledge is, in theory, available to anyone with an internet connection. But despite this apparent abundance, there is a deep and troubling problem at the heart of our digital world: the systems we rely on for knowledge are not designed to give us what is true. They are designed to give us what is popular. This single flaw has led to the quiet dominance of falsehood across the internet, and it has profound consequences for education, religion, public discourse, and the very idea of truth itself.
Search engines such as Google and AI systems like ChatGPT do not fundamentally operate as instruments of truth-seeking. Rather, they are engines of pattern recognition and relevance scoring, designed to serve users content that appears statistically likely to satisfy their queries. For a search engine, this means results that receive the most clicks, shares, backlinks, and engagement are treated as authoritative. For a large language model, it means that the most frequently repeated and co-occurring phrases in its training data are considered the most reliable responses. In both cases, truth is not the goal. Engagement is. The goal is to deliver something that looks like the right answer, feels like the right answer, or matches the user’s expectations — whether or not it actually reflects historical fact or metaphysical reality.
This creates a subtle but powerful bias: the truth is filtered through the lens of popularity. If an idea is well-marketed, if it appeals to the tastes of the moment, or if it has been repeated enough times across various platforms, it becomes “true” in the eyes of the algorithm. On the other hand, if a truth is obscure, difficult, or unpopular — even if it has stood the test of time for centuries — it may be entirely invisible in a digital search. The result is a distortion of knowledge, in which the most viewed, most liked, and most quoted content dominates public understanding, regardless of its accuracy.
Take, for example, the subject of classical education. If one asks ChatGPT or Google to explain what classical education is, the typical response includes references to Dorothy Sayers’ essay “The Lost Tools of Learning,” Susan Wise Bauer’s book “The Well-Trained Mind,” and organizations like the CiRCE Institute, Classical Conversations, and Memoria Press. These names and ideas dominate the modern discourse on classical education. But none of them represent the historical classical tradition. Dorothy Sayers was a 20th-century novelist who wrote a speculative essay on education. Susan Wise Bauer is a Protestant homeschooling advocate. The CiRCE Institute is a modern think tank. None of these sources reflect the true, Catholic, historical structure of classical education — the education of Aristotle, Cicero, Boethius, Augustine, Thomas Aquinas, and the Church Fathers. And yet, because they are widely circulated and referenced, they become the standard definition of classical education online.
Meanwhile, authentic programs like the Classical Liberal Arts Academy, which offer a full recovery of the actual curriculum used for over two thousand years, including the trivium, quadrivium, philosophy, theology, and the teachings of the Church, are not mentioned at all. Not because they are false. Quite the opposite. But because they are not sufficiently repeated, linked, or quoted in the mainstream digital ecosystem. They are drowned out by the noise of more popular but less accurate voices. This is not just an unfortunate oversight. It is a symptom of a much larger disease in the way we acquire and validate knowledge in the modern world.
This pattern is not unique to education. It repeats itself in theology, history, science, and politics. Wherever the algorithm governs what we see, popularity governs what we believe. And because most people do not go beyond the first page of search results, or the first paragraph of an AI answer, their understanding is shaped almost entirely by the consensus of the crowd. In an age that celebrates individual thinking and free inquiry, we have quietly surrendered the authority of truth to the invisible hand of digital popularity.
This is a radical departure from the way knowledge was once pursued. In the classical and medieval traditions, knowledge was not measured by its popularity but by its conformity to reality. Philosophers labored through dialectical reasoning to identify and refute error. The Church safeguarded divine revelation and tested doctrines against the rule of faith. Scholars and teachers submitted their ideas to centuries of refinement, critique, and synthesis. The truth was treated as something objective, independent of opinion, and worth suffering for. But today, truth is treated as something subjective, consensus-based, and only valuable if it pleases the algorithm.
The internet has made this worse, not better. While it provides access to vast amounts of information, it does not help us evaluate the quality of that information. Worse, it often rewards the opposite of quality. Shallow ideas that can be summarized in a tweet or catchy headline spread more easily than rigorous explanations that require sustained attention. Controversial takes generate more clicks than careful distinctions. Emotional appeals outperform rational arguments. The result is a flood of viral content that looks like knowledge but lacks the substance of true understanding. It is fast food for the intellect — satisfying in the moment but nutritionally empty in the long term.
This creates a world in which falsehood can not only survive but thrive. It doesn’t matter if an idea is false, if it serves a narrative, if it confirms a bias, or if it sells a product. It will be rewarded by the system as long as it performs well. Even well-intentioned content creators are tempted to cater to the algorithm by simplifying, exaggerating, or flattening the truth. Over time, the very structure of online discourse changes to match what the machine prefers — and the machine prefers what is popular.
Artificial intelligence adds another layer to this problem. While search engines rely on user behavior to determine relevance, AI systems like ChatGPT are trained on enormous datasets that reflect the totality of online discourse, including books, blogs, academic papers, news articles, and forum posts. This means the model is shaped by the most common patterns of language and association in its training data. If a particular error is repeated often enough, it becomes encoded in the model’s statistical understanding of how to answer related questions. It does not have a concept of “truth” in the way a human being does. It has a concept of “probable response.” And so, falsehood becomes the default — not because the AI wants to lie, but because it has no access to metaphysical standards of truth. It only knows what it has seen, and what it has seen is what we have most frequently said.
This is why a person asking about the meaning of classical education, the history of Catholic liturgy, the origins of Western philosophy, or the true nature of scientific method is likely to receive a well-formatted, confident, and completely shallow answer. The system has no commitment to truth. It has a commitment to coherence and user satisfaction. It tells people what they expect to hear, not what they need to know. And because it sounds fluent, most users will never question it.
The implications of this are enormous. If we are forming students, citizens, and Christians in a world governed by these systems, we must be aware that they are not neutral tools. They are embedded with assumptions, biases, and patterns that reflect the values of the crowd. If the crowd has abandoned truth in favor of relevance, novelty, or pleasure, then the tools we use will reflect and reinforce that abandonment. And if we allow these tools to shape our thinking without discipline, we will be slowly conformed to their image — trading wisdom for convenience, and truth for consensus.
This dynamic mirrors the effects of democracy in the political sphere, where the will of the majority overrides the guidance of truth and virtue. In a system where each opinion counts equally, regardless of wisdom, experience, or moral integrity, the majority inevitably steers policy, culture, and law toward what is most popular or emotionally persuasive rather than what is right. Since the majority is almost always composed of individuals lacking deep education, philosophical discipline, or theological formation, the direction they choose tends to reflect shallow desires, not enduring principles. Just as in the digital world algorithms elevate what is most clicked rather than what is most accurate, democratic societies often elevate what is most agreeable rather than what is most just. In both cases, the result is the triumph of quantity over quality, sentiment over reason, and fashion over truth. This is the great deluding influence of modern society.
What Can Be Done?
The question, then, is what can be done. The first and most important step is to recognize the problem. We must stop pretending that digital systems are reliable guides to truth. They are not. They are efficient delivery systems for content, not discerning judges of reality. I use AI tools to help me compose articles (like this) to allow me to delegate the basic work of “arrangement” and devote my time and energy to the more important work of “invention” (See the Five Canons of Rhetoric). Once we accept that, we can begin to use them appropriately. We can learn to direct these resources by carefully prepared prompts and instructions, then question results, verify information, and seek out reliable sources that are not popular but are proven. We can teach our children and students to think with the mind of the Church, the philosophers, and the saints — not the trends of the day. If we use the technology responsibly, it can be incredibly useful–and we must do so.
Second, we must rebuild the infrastructure of truth in the digital world. That means writing, publishing, and sharing content that is grounded in real history, sound philosophy, and faithful theology. It means creating alternative tools — AI models, search indexes, archives, and platforms — that are not governed by popularity but by principle. It means rewarding depth over clicks, clarity over controversy, and faithfulness over virality.
Third, we must embrace our responsibility as guardians of tradition. The truths we hold — whether they are philosophical, scientific, or theological — have been handed down to us at great cost. They are not merely opinions. They are the foundation of civilization. If we do not protect them, teach them, and transmit them, they will be buried beneath the noise of the age. But if we do, we can preserve a remnant of wisdom that will outlast the digital fog.
The age of popularity-based knowledge is not sustainable. Sooner or later, a society that prefers what is popular over what is true will collapse under the weight of its own illusions. Truth does not go away. It can be ignored, distorted, and forgotten for a time. But it always returns, because it corresponds to what is. Our task, then, is not to despair at the dominance of falsehood, but to remain faithful to the truth. To teach it. To live it. And to build anew, even in the ruins of modern information.
If we do that, then no matter how loud the lies become, we will have done our part to keep the light of truth visible. Truth is not preserved by crowds or trends, but by those willing to study it, teach it, and defend it. In an age where falsehood spreads effortlessly through systems built on popularity, the preservation of truth requires deliberate and disciplined effort. It requires the commitment of individuals and institutions who are willing to resist the current, speak with clarity, and remain anchored in what is objectively real. This is the task before us — and it will not be fulfilled by accident.
Mr. William C. Michael, O.P.
Headmaster
Classical Liberal Arts Academy
