We live in an era of imminent explainability. We can explain how pathogens spread disease, what factors produce mental illness, and how fermentation makes dough rise. Once attributed to supernatural, divine, and moral forces, we can now elucidate these processes through clear scientific principles. Painstaking experimentation, education, and scholarly exchange have enabled humans to learn so much. And now, with the advent of generative artificial intelligence (AI), any one of us can explain even more, without years of studying or hours spent thinking. That might not be a good thing.
When German sociologist Max Weber lamented over a century ago that the scientific, rational, and technical explanation ushered in by the Enlightenment contributed to the world’s “disenchantment,” he could have scarcely imagined the disenchanting capabilities of generative AI. This technology has the comprehensive capacity to promptly answer any question (rightly or wrongly), explain complex phenomena, and perform many functions exceeding human capabilities. Truly, talking to large language models (LLMs) like ChatGPT, Claude, or Gemini can feel magical. But do these conversations also deprive us of opportunities for awe, wonder, and pure amazement?
In his Vocation Lectures, Weber bemoaned that while scientific “progress” (a term he used disparagingly) offers explanation, it simultaneously robs us of meaning: “Science gives us answers to the question of what we must do if we wish to master life technically. It gives no answer to the question of whether we should and whether it ultimately makes sense to do so.”
That is, science tells us what is, not what ought to be.
And fixating on the explanations for our present reality may undermine our ability to dream up new ones.
Given generative AI’s similar role as an explanation generator—an analysis of 130,000 users demonstrated that people primarily use ChatGPT for answering questions (rather than for performing tasks or expressing themselves)—we might ask whether such immediate access to epistemic closure has dampened our sense of meaning, hastening a global malaise. The term Weber used to describe this state, “Entzauberung,” translates to de-magic-ing. What is the cost of losing a sense of magic in the world? And how might we bring it back?
The Danger of Certainty
In many ways, generative AI’s rise represents an accelerated evolution of preceding tools that offered immediate answers (like Google) and facts (like Wikipedia) and that dramatically reduced the discomfort of “not knowing” over prolonged periods of time. “What was the name of that actor who made the sound effects with his voice in the Police Academy movies?” “How can I hear that Rihanna song again?” “Where is the best place to eat gluten-free pizza nearby my dentist’s office?” “What driving route do I take to get there?”
These are questions that can be answered within seconds. But the questions that people implore LLMs to answer involve much broader issues, such as medical prognoses, mental health issues, professional challenges, and existential concerns. Notably, the answers that LLMs provide to these questions often seem to satisfy people’s quest for resolution and comprehension, irrespective of the actual wisdom or accuracy contained in them.
LLMs also differ from prior recent technologies in other important ways that likely hamper awareness of one’s intellectual limitations. Unlike traditional web search that often involves an iterative back-and-forth process of pursuing answers to questions, interactions with LLMs tend to be brief. An analysis of over 13,000 ChatGPT sessions showed that “users send an average of just 1.7 messages per conversation.” Relatedly, using LLMs as recommendation systems steers people toward exploring a less diverse and distinctive set of choices than what they would otherwise discover. Compared to prior internet tools, LLMs are increasingly likely to curtail inquiry rather than spur further intellectual exploration.
People critique generative AI for fabricating information, generating slop, and forming unhealthy relationships with users. These are issues that engineers, programmers, and AI companies are currently working to fix. Yet, the sense of certainty that LLMs provide contributes to something potentially more dangerous—a decline of intellectual humility. Recognizing the fallibility of our own beliefs and knowledge has considerable societal and personal benefits that we may fail to reap if we over-rely on AI. As LLMs ensure we rarely have to experience being wrong, uncertain, or baffled, we mistakenly feel we have no reason to doubt our own beliefs. These findings make a case for actively encouraging people to recognize the limits of what they know, but this poses a challenge, given our general disdain for feeling wrong, dumb, or incompetent.
The Allure of Magic
I would like to propose a solution that reminds people of their ignorance without threatening their egos: exposure to magic—that is, events that appear to defy physical or scientific causation.
Indeed, encountering events that challenge our understanding of the world produces a sense of awe that can increase humility and even reduce ideological extremity. And fortunately, people appear primed for such a return to enchantment. Whereas my own research has demonstrated that, consistent with what Weber would have predicted, the rise of artificial intelligence has precipitated a decline in religiosity, people have also become increasingly inclined toward alternative spiritual practices ranging from tarot cards and astrology to healing crystals and even witchcraft.
Beyond this propensity toward magic in the form of superstition, however, we can also conceive of magic as a way of engaging with mystery and the unknowable.
Re-enchanting Your Life
Consider some practical steps for cultivating a reenchanted world derived from principles in both secular and religious contexts.
First, we can resist the modern obsession with optimization and instead practice particular activities for their own sake. Activities like taking a walk without a destination, talking to strangers, approaching the day without a to-do list, or intentionally reading books that interest us build more serendipity into our lives and critically remind us of the indeterminacy of the world. This embrace of non-instrumentality for me is best captured by French industrial designer Philippe Starck’s Hot Bertaa water kettle, an aesthetic feat of postmodern design that also fundamentally does not work, offering no way to measure the water level and often scalding its users. As a tea kettle, it fails spectacularly, but as a sculpture, it is magnificent in its asymmetric and minimalist nature, reminding me that non-instrumentality is valuable in its own right.
Second, we can actively seek out experiences that defy simple explanation, from visiting a contemporary art museum to hiking amongst the natural beauty of The Painted Desert, to attending a symphony or a rave. And although spending time marveling at the cosmos is a millennia-tested strategy for experiencing wonder, artificial intelligence may have a complementary role to play, as it has begun to design architecture and generate scientific hypotheses in ways its creators do not even fully understand. Whether we prefer the majesty of the natural world or the miracles of modernity, we will feel more fulfilled when we seek situations that force us to wrestle with life’s mysteries rather than simply master them.
Third, we can reduce our fear of uncertainty, a practice that emerges in both sacred and secular traditions throughout history. Ancient Greek Philosophy championed the virtue of “Socratic ignorance” or admitting the limits of one’s knowledge. Zen Buddhism teaches the importance of Sho Shin, or “beginner’s mind,” which also involves freeing oneself from what one knows to approach the world with openness. Laozi, founder of Taoism, describes “not-knowing as true knowledge.” Other traditions tie the importance of embracing one’s intellectual limitations to interactions with the divine, with elements of Islamic thought, Jewish scholarship, and Christian mysticism all emphasizing that human reason and intellect are insufficient for understanding God. Across these traditions, treating “not knowing” as an essential virtue becomes liberating, helping us to accept the limits of our own minds and bolstering our humility.
Taken together, re-enchanting your life is a formidable challenge, but I can suggest one exercise that has helped me on this journey: These days, I simply say “I don’t know” more often. Whether a student asks me to solve a difficult moral dilemma in my lecture course on ethics, or my 7-year-old daughter poses a question from the backseat of the car (“Dad, what was the first word ever spoken on Earth?”), I have found incredible liberation in responding that I simply don’t know. The brief reminder that I have no idea how, at some point, humans collectively decided on a moral code or invented language feels downright…magical.
Despite its association with the irrational, the occult, or scam-laden parlor tricks, experiencing magic in ways described here might produce a more accurate understanding of the world as we embrace its true complexity, echoing Voltaire’s sentiment that “Doubt is not a pleasant condition, but certainty is an absurd one.” By embracing mystery in the face of technological mastery, we can regain intellectual humility that makes us more open to understanding each other.
