In Goethe's 1797 poem “The Sorcerer's Apprentice,” the apprentice, alone in his master's house, wants to multiply himself to shirk his work. He enchants a broom to fetch water from the well so he will not have to. The broom, lacking any cognitive abilities or self-awareness, executes the command relentlessly.
When the apprentice tries to stop it, he discovers he does not know the deactivation spell. He chops the broom in two, and now there are two brooms mindlessly fetching water. The house floods.
Delegation to a thing without cognitive or moral understanding can’t multiply the self, but it does amplify a part of the self that gave the order.
Despite adaptations of “The Sorcerer's Apprentice” in popular culture, Mark Zuckerberg, Jack Dorsey, and many others seem to have missed the lesson. They have their own fantasies about reducing their workload and extending control.
Mark Zuckerberg believes an AI representation of himself can be his “broom,” and reportedly, Meta is developing “a photorealistic, three-dimensional AI avatar.” Trained on Zuckerberg’s mannerisms, remarks, and corporate strategies, this avatar is designed to interact with Meta employees, answering questions and offering managerial guidance on his behalf.
Similarly, Jack Dorsey, CEO of Block, laid off a massive portion of his workforce and expressed a desire to use an AI “intelligence layer” to completely flatten the management hierarchy. Dorsey envisions a scenario in which middle management is eliminated, and all 6,000 employees report directly to him via an AI intermediary.
Flattening an organization may be an admirable goal, yielding adaptive outcomes. But will AI get you there? (And is it even necessary? Decades before AI, W.L. Gore and Associates, of Gore-Tex fame, began reaping the benefits of a hierarchy-free organization, and they are still going strong.)
Can Machines Be Conscious
The work of Federico Faggin, the pioneering physicist and inventor of the first commercial microprocessor, provides a scientific and philosophical rebuttal to the idea that AI could ever serve as a genuine leader.
When Zuckerberg and Dorsey attempt to replace human managers with an “intelligence layer,” they are replacing human understanding with blind algorithmic execution. Faggin explains in his book Irreducible “that unconscious matter cannot produce consciousness.”
Faggin argues that human intelligence is superior because it depends on genuine comprehension, a “non-algorithmic property of consciousness that computers can never possess.” Machines may process data faster, but, in his view, “true intelligence is the ability to comprehend,” which means “to go beyond the immediate meaning of symbols” by grasping their broader context. True intelligence remains beyond the reach of a machine.
Placing AI in a leadership position means handing authority over to a mechanism that can only mimic intelligence but fundamentally lacks the intuition, empathy, and ethical judgment required to lead.
Faggin writes, “Meaning, understanding, and free-will decisions do not exist within a computer.” Faggin notes that the Italian word conoscere means the knowing that rests on “experience and comprehension,” while sapere describes “the knowing by heart of someone who repeats what he does not understand.” Computers do not comprehend. Self-awareness cannot arise out of matter.
No consciousness is present in Zuckerberg's avatar on a video call with an employee. It does not understand the employee's question; it will pattern-match a response. When the employee leaves the call believing she has “talked to Mark,” she has in fact talked to an echo of Zuckerberg’s prior understanding.
Faggin's point is not a romantic, goal-line defense of humanity; it is rooted in scientific rigor. Only the employee receiving AI “guidance” is capable of understanding. The conversation is asymmetrical; one party is conscious; the other is reciting.
An executive's value to the firm lies in the entrepreneurial capacity to recognize when a situation does not fit the existing pattern, which is precisely the part that cannot be modeled. What the avatar reliably reproduces is the part of the CEO that is already cliché, the part that employees endure in boring meetings.
Another philosopher/computer scientist, Bernardo Kastrup, in his book Analytic Idealism in a Nutshell, writes, “AI chatbots are not intelligent in any way remotely akin to how you and I are intelligent.” They are, he says, “merely natural language interfaces to search engines” and “in fact have no understanding of what they are saying; none whatsoever.”
Kastrup warns of dangers, both psychological and spiritual. Because the output arrives in conversational prose, “we no longer see the sources of the information being conveyed to us,” and we begin to treat the answers “as if they were truths pronounced by an all-knowing oracle who understands what it is saying.” He warns we are losing “our ability to be critical of [AI’s] answers.”
In The Daimon and the Soul of the West, Kastrup observes that many have convinced themselves that they are masters of reality: “We tell ourselves that we can be whoever we like to be; that we can not only do as we want, but—absurdly—want as we want. Masters of the self we are—or so we want to believe.” Deluded CEOs attempt to play God within their corporate fiefdoms, seeking to control thousands of people simultaneously without the messy, unpredictable friction of actual human relationships.
So, what does this have to do with you and me? Why might so many of us voluntarily surrender our agency to digital tyranny?
The Threat to Civilization
Rod Dreher, the author of the excellent Live Not by Lies, has been sounding the alarm about where this reality-defying hubris may lead society.
In his chilling essay, “AI And Weimar America,” Dreher argues that the same misunderstanding that leads CEOs to deploy AI intermediaries is pushing human civilization toward a catastrophic surrender of its own agency.
This voluntary surrender of power is Dreher’s primary concern. He asks whether we are sleepwalking into creating an AI “neo-Caesar” that will rule over us, not because it seized power but because we willingly handed over the keys to civilization for the sake of convenience and efficiency.
Dreher draws a dark parallel between our current era and the “psychosocial” conditions of pre-Nazi Germany. Having written about Weimar Germany here and here, I find Dreher’s parallel a yellow flag worth our consideration.
He notes that the youth of the Weimar Republic were “socially atomized, adrift from their own authoritative traditions and institutions, and stressed maximally by the chaos that they had been living through.”
Just as dictators of the past offered relief from misery in exchange for absolute surrender, the new technological rulers will offer a digital savior. Dreher predicts that we will “surrender our agency and humanity to some form of managerial technocracy that uses superintelligent AI to win our allegiance by promising us stability and happiness in exchange for our liberty.” Those who don’t accept the deal “will be dealt with.”
Dreher's warning is that this digital tyranny will be embraced by a public eager to escape the responsibilities of freedom. Many people will welcome the AI Caesar, doing “anything to relieve them from the psychic burden of living.” They will prefer to live inside a “pleasing, comforting lie,” much like a lonely person finding happiness with an AI lover they know is fake, rather than face the harsh truths of reality.
However, the core existential threat Dreher identifies is not a violent, cinematic machine uprising. The true terror lies in human complicity and voluntary surrender. Are we, Dreher wonders, on a trajectory to intentionally hand over the foundational pillars of civilization to AI systems?
AI’s growing fluency is precisely the danger, because it widens the gap between how conscious it seems and how absent consciousness is.
No matter how much AI flatters you, it doesn’t care about you. It doesn’t appreciate you or understand you. It never will.
A broom cannot know what a house or civilization is for. Neither can an AI Caesar. Goethe's apprentice did not know the deactivation spell to end the flood. We know it perfectly: respect dispersed intelligence, accept the responsibilities of freedom, and govern ourselves. Some have simply decided they would rather drown than govern themselves.
