In a sweltering Renault in Marseille last summer, a simple wrong turn became a profound metaphor for our technological age. A navigation app, Waze, overruled a friend's local knowledge, leading directly into a construction site dead-end. This minor frustration encapsulates a defining modern dilemma: in an era where artificial intelligence (AI) touches everything, who do we trust more—human instinct or the machine?
From Kant to ChatGPT: A History of Intellectual Emancipation
The German philosopher Immanuel Kant defined the 18th-century Enlightenment as humanity's emergence from a "self-imposed immaturity." This state, he argued, was the inability to use one's own understanding without guidance from another. For centuries, that guiding "other" was often a priest, a monarch, or a feudal lord—figures claiming divine authority over natural and social order.
Kant believed humans always possessed the capacity for reason but lacked the confidence to wield it. The revolutions in America and France heralded a new dawn where reason would supplant blind faith. The rallying cry, "Sapere aude!" or "Have courage to use your own understanding!" championed the human mind as the engine of moral and material progress.
Fast forward two and a half centuries, and one must ask: are we quietly slipping back into a new form of immaturity? An app suggesting a route is trivial, but AI's rapid ascent positions it as a potential new "other"—a silent, algorithmic authority guiding our choices, from the mundane to the monumental.
The Allure of the Algorithm: Convenience at the Cost of Cognition
The statistics are startling. ChatGPT launched just three years ago, yet an April global survey found 82% of respondents had used AI in the prior six months. People now consult machines on deeply personal matters, from ending relationships to voting decisions. OpenAI data reveals 73% of user prompts concern non-work topics, with writing assistance among the most common uses.
This dependence has a cognitive cost. A study from the Massachusetts Institute of Technology (MIT) used electroencephalography (EEG) to monitor brain activity in essay writers. Those with AI access showed the lowest cognitive engagement and had difficulty accurately quoting their own work. Over time, they grew lazier, copying entire text blocks. The study's scale was limited, but its findings echo Kant's warning about the ease of immaturity born from "laziness and cowardice."
AI's appeal is undeniable: it saves time, effort, and offers a novel way to offload responsibility. The psychoanalyst Erich Fromm, in his 1941 book Escape from Freedom, argued that fascism rose partly because people traded freedom for the certainty of subordination. AI presents a digital-age version of this surrender, relieving us of the burden to think and decide.
The Black Box of Faith: When Reason is Replaced by Reliance
The core problem lies in AI's opacity. It generates knowledge without necessarily fostering human understanding. Even its programmers often cannot fully explain its reasoning, and we lack clear criteria to verify its conclusions. When we follow AI's advice, we are not being guided by transparent reason; we are operating on faith. The principle "in dubio pro machina"—when in doubt, trust the machine—threatens to become our new credo.
This is not to dismiss AI's formidable benefits. It can accelerate drug discovery, liberate us from tedious "bullshit jobs," and handle complex tasks like tax returns. However, Kant and his peers did not advocate for reason over faith merely for efficiency gains or more leisure time. Critical thinking was, and is, a practice of freedom and human emancipation.
Human thought is messy and error-prone, but it necessitates debate, doubt, and testing ideas. This process builds individual and collective confidence. For Kant, exercising reason was about enabling people to become agents of their own lives and resist domination, fostering a moral community grounded in shared debate rather than blind belief.
The central challenge of the 21st century is thus clear: how do we harness AI's superhuman capabilities without eroding the human reasoning that is the cornerstone of the Enlightenment and liberal democracy itself? It is a question of profound importance—and one we would be wise not to delegate to the machine.
