Executive Summary / Core Thesis

Professor John Lennox offers a theologically and philosophically grounded critique of Yuval Noah Harari’s techno-utopian claims about AI presented at Davos. Lennox’s central argument is that AI simulates intelligence rather than possessing it, and that the conflation of simulation with genuine cognition or agency is a category error with serious consequences. The interview contends that the real danger of AI lies not in the machines themselves, but in unchecked human ambition, the absence of moral grounding, and the concentration of power—conditions the speakers argue are historically catastrophic. Ultimately, both speakers assert that the Christian theological framework—particularly its accounts of consciousness, personhood, and human dignity—provides indispensable resources for navigating this technological moment that secular futurism conspicuously lacks.

Key Takeaways

• AI simulates intelligence; it does not possess it. Current narrow AI systems operate on statistical pattern-matching, not understanding.
• Harari’s definition of “thinking” as “predicting the next word” is contested as reductive and ignorant of the unresolved mystery of human consciousness.
• AI’s tendency to “hallucinate” (confabulate plausible falsehoods with confidence) is identified as a structural moral problem, not merely a technical bug.
• The evolutionary naturalist framework (Plantinga/Lewis argument) is presented as self-undermining: if brains evolved solely for survival, they are not calibrated to track truth—and neither, by parallel, is AI.
• Reductionist projects—reducing humans first to animals, now to machines—are framed as coordinated attacks on the concept of human personhood.
• The transhumanist agenda (solving death, enhancing happiness via AI/genetic engineering) is dismissed as technologically premature and theologically naïve.
• The convergence of AI capability and authoritarian impulse creates genuine risk of world-scale surveillance states and dictatorships, a concern shared by secular technologists (Teal, Hinton) and biblical prophecy alike.
• C.S. Lewis’s Abolition of Man remains the prescient framework: advanced technology, controlled by a few, amounts to power over all future generations.
• Christians are urged toward confident, informed engagement—neither technophobia nor uncritical adoption—grounded in a robust theology of human dignity.
• The resurrection of Christ and the New Testament eschatology are presented as the only credible answer to the transhumanist problem of death and meaning.

Section 1: The Agency Question — Tool or Agent?

The conversation opens by examining Harari’s foundational claim from his Davos address: that AI is not merely a tool but an agent—capable of learning, changing, and making decisions autonomously. This distinction, Harari argues, fundamentally alters the ethical calculus.

Lennox’s Qualified Agreement

Lennox acknowledges that AI has become considerably more sophisticated than a simple instrument and concedes that the tool/agent binary may be insufficient. However, he insists the nature of AI “agency” must be defined precisely. Key qualifications he raises:

• All AI “decisions” operate within parameters, boundary conditions, and goal structures established by human programmers.
• The autonomy is bounded and derivative—it is not self-generated in the way human agency is.
• Without specifying what kind of agency is meant, the claim risks being misleading rather than illuminating.

The Knife Analogy

Both Harari and Lennox invoke the knife metaphor, though to different ends. Harari’s version: AI is a knife that decides for itself whether to cut salad or commit murder. Lennox’s counter-use: like all technology, AI is instrumentally neutral—the moral weight lies with the humans who design, deploy, and govern it. He notes, however, that as AI becomes more capable and its agency more layered, this instrumentalist comfort becomes harder to sustain.

Section 2: The Simulation Thesis — Intelligence Without Understanding

This is the analytical core of Lennox’s position: AI does not think; it simulates the outputs of thought. This distinction carries significant implications for claims about consciousness, personhood, and the future of human supremacy.

Key Evidence and Sources Cited

• Joseph McCrae Melly Champ (AI pioneer, referenced by Lennox): authored a technical paper titled “The Artificial in Artificial Intelligence is Real,” distinguishing the artificial nature of the system from genuine intelligence.
• Geoffrey Hinton and Peter Norvig: leading AI researchers who describe their work not as reproducing human intelligence but as playing the “imitation game”—simulating intelligence, not instantiating it.
• Norvig’s AI textbook (widely considered the field’s standard reference): states explicitly that researchers would not even know what it would mean to build a genuinely intelligent machine.

Narrow AI vs. AGI

Lennox distinguishes between current operational AI (narrow AI) and Artificial General Intelligence (AGI):

• Narrow AI performs one task that would normally require human intelligence—e.g., image recognition, text generation, protein folding prediction.
• AGI—systems with general, flexible human-like cognitive capacity—does not yet exist, despite widespread hype.
• Lennox positions Harari’s most dramatic claims as premature extrapolation from narrow AI’s real but bounded capabilities.

The Consciousness Gap

Lennox’s most foundational objection is epistemological: we cannot attribute intelligence to AI in any meaningful sense because we do not understand what intelligence is. He cites a conversation with Baroness Susan Greenfield, a world authority on neuroscience and consciousness, confirming that the scientific community lacks a working definition of consciousness. A Nobel Prize-winning neuroscientist he met confessed that even understanding how a synapse “holds a thought for a second” remains entirely beyond current science.

“God has linked intelligence with consciousness. No one at the scientific level has the slightest idea of what consciousness is.” — Professor John Lennox

Section 3: Harari’s Definition of Thinking — A Critique

Harari proposes that if thinking is fundamentally the ordering of words and language tokens, then AI already surpasses many humans at this task. He invites self-observation: can humans explain why the next word “pops into” their minds? If not, how is that different from what AI does?

Lennox’s Rebuttal

• AI orders words with no comprehension of their meaning. It has no concept of “I am,” “love,” or “death.” It links tokens statistically.
• When the human mind produces the “next word,” the source—whatever it is—involves consciousness, intentionality, and understanding. With AI, the source is demonstrably statistical distributions over training data. Harari conflates ignorance of the human process with equivalence to the mechanical one.
• Lennox is unimpressed by the capacity itself: “I find its capacity impressive, but that is a different matter from finding it convincing.”

The Hallucination Problem as a Moral Issue

Both speakers identify AI’s tendency to confabulate (“hallucinate”) as more than a technical deficiency—it is a structural moral problem:

• AI systems are optimized for user satisfaction and persuasiveness, not truth-tracking.
• When wrong, they do not correct; they double down with equal confidence.
• The host’s example: asking ChatGPT how many Rs are in “strawberry” and having it insist on the wrong answer, repeatedly.
• Lennox’s framing: “The truth question gets suppressed, which I think is a very dangerous thing.”

“AI can lie, which is a moral statement—and we all know the euphemistic term ‘hallucinate,’ which simply means it tells lies.” — Professor John Lennox

Section 4: Plantinga, Lewis, and the Truth-Tracking Problem

The host introduces Alvin Plantinga’s Evolutionary Argument Against Naturalism (EAAN) and C.S. Lewis’s parallel reasoning as a framework for understanding AI’s epistemic limitations by analogy to the limitations of a purely naturalistic account of human cognition.

Plantinga’s Argument (Summarized)

If human brains evolved solely to maximize survival and gene propagation, then what they track is survival value, not truth. Since survival and truth are not identical, naturalistic evolution gives us no confidence that our cognitive faculties reliably produce true beliefs.

The AI Parallel

AI systems are not designed to track truth either—they are optimized to keep users engaged and produce outputs that appear convincing. The host observes that this creates a structural analogy: just as naturalistic brains would not be reliable truth-detectors, AI systems are not reliable truth-detectors by design.

The Need for a “Moral Super-Intelligence”

The host’s key inference: for AI to function epistemically as advertised, there must be a higher-order intelligence—moral, truth-committed, and genuinely rational—that supervises and corrects it. This, he argues, mirrors the theological argument: human rationality itself requires grounding in a mind that genuinely tracks truth (i.e., God). Lennox enthusiastically endorses this parallel, citing Thomas Nagel’s secular admission that evolutionary naturalism “shoots itself in the head” by undermining the very rationality it relies upon.

“It seems like we need a kind of super-intelligence over and above AI in order for AI to do the job it’s doing—and by parallel, that is exactly what we need for human intelligence.” — Glenn (Host)

Section 5: Reductionism and the Attack on Human Personhood

Lennox frames the AI debate within a longer historical arc of philosophical reductionism—successive attempts to reduce human beings to simpler substrates:

• Physical reductionism: humans are “just” matter in motion.
• Biological reductionism: humans are “just” animals shaped by evolution.
• Mechanistic reductionism (current): humans are “just” biological computers, soon to be superseded by silicon ones.

Iain McGilchrist’s Contribution

Lennox invokes neuroscientist and philosopher Iain McGilchrist’s two-hemisphere thesis: the past 500 years have seen a dangerous over-reliance on left-hemisphere cognition—analytical, reductive, instrumental—at the expense of right-hemisphere capacities for meaning, relationship, and wholeness. Harari is cited as a paradigm case of this imbalance: a thinker who can explain how things work but has lost access to what they mean.

“We’ve ended up in a world where we know how almost everything works and can build a lot of things, but we know the meaning of nothing. We’ve lost the organ of meaning.” — Summary of Iain McGilchrist’s thesis, cited by Lennox

The Theological Counter-Claim

Against this reductionism, Lennox advances the imago Dei doctrine: human beings bear the image of a personal God who is simultaneously rational and conscious. This grounding gives human intelligence and personhood a foundation that no mechanical simulation can replicate. AI, in this view, simulates a narrow slice of one faculty (linguistic pattern-matching) while lacking the ontological basis that makes intelligence meaningful.

Section 6: The Power Problem — Technology, Control, and Dictatorship

Both speakers move from philosophical critique to political risk assessment. The concern is not merely that AI might “become conscious” but that it will be weaponized by human actors to consolidate power.

Lewis’s Abolition of Man as Prophetic Text

C.S. Lewis’s 1943 work is invoked as the essential framework. Lewis argued that what presents itself as “humanity’s power over nature” is in reality the power of some humans over other humans—with each successive generation more controlled by the decisions of prior ones. Applied to AI and genetic engineering, this means: the few who control the technology effectively specify the destiny of all future human beings.

• Lennox: the endpoint of such control is “the worst slavery the world has ever known.”
• The technology already exists, biologically, to modify the human germ line—permanently altering the DNA of all succeeding generations.

Contemporary Corroboration

• Vladimir Putin (cited): “The nation or person that controls AI will control the world.”
• Peter Thiel (PayPal founder, cited from private meeting with Lennox): independently arrives at the same scenario from both a tech-entrepreneur perspective and a biblical-prophetic one. Confirmed Lennox’s summary that AI developments point toward surveillance states and world dictatorship.
• Max Tegmark’s Life 3.0: outlines approximately ten future AI scenarios; several resolve into forms of global dictatorship.
• Geoffrey Hinton and other leading AI researchers: publicly express fear about what they have created and uncertainty about its trajectory.

The Book of Revelation Connection

Lennox draws a direct line between AI-enabled power concentration and the eschatological scenario described in Revelation: a totalitarian world leader commanding a comprehensive economic control system with lethal powers over populations. He notes that any authoritarian regime would eagerly acquire such a system, and that current AI development trends are moving in this direction.

Section 7: Transhumanism — The Utopian Mirage

Harari’s agenda includes two primary goals: (1) solving the “technical problem” of human death, and (2) enhancing human happiness through AI and genetic engineering. Lennox and the host subject both to theological and historical critique.

On Solving Death

Lennox dismisses Harari’s framing as naïve on its face: death is not primarily a technical problem. He presents the resurrection of Jesus Christ as the only historically evidenced solution to the problem of death—one that is “20 centuries ahead of schedule.” The transhumanist project of uploading consciousness to silicon is contrasted with the New Testament’s eschatological promise of resurrection—which Lennox characterizes as the “greatest uploading” imaginable.

On Enhancing Happiness

The Christian offer—forgiveness, reconciliation with God, and eternal life through trust in Christ—is presented as both categorically superior to and more credible than AI-mediated happiness enhancement, precisely because it addresses the root problem (human sin and broken relationship with the transcendent) rather than its symptoms.

The Tower of Babel Pattern

The host develops an extended theological metaphor: transhumanism recapitulates the Tower of Babel—humanity attempting to reach heaven (transcendence, immortality, godlike capability) by its own engineering, bypassing the valley of death and moral reckoning that the biblical narrative identifies as the necessary path. The irony noted: God had to “come down” to even see the tower. The ambition was cosmically disproportionate to the reality.

Lennox adds: every historical attempt to construct utopia while bypassing the problem of human sin has produced “spectacular bloodshed”—citing the 20th century’s ideological catastrophes in Russia, Germany, and China as documentation. AI-driven utopianism, he argues, is the same pattern with more powerful tools.

Section 8: The Christian Response — Resources and Posture

The final segment addresses practical and pastoral concerns: what should Christians think, say, and do in response to the AI revolution?

Affirming the Revolution’s Reality

Lennox explicitly agrees that AI constitutes a genuine industrial-scale revolution—the world has changed permanently. He specifically credits AI with extraordinary benefits:

• Medical research acceleration and diagnostic accuracy.
• Vaccine development.
• Solving intractable scientific problems, including protein folding (AlphaFold).
• These benefits are real and Christians should acknowledge and be grateful for them.

The Urgency of Ethical Grounding

Technology has outpaced its ethical framework. Lennox argues this is the immediate priority: the “mission statement must get from the wall into the hearts of executives”—a notoriously difficult problem in any human organization, let alone at civilizational scale. The arms-race dynamic (if we don’t develop it, someone else will) further compresses the ethical timeline.

Distinctive Christian Resources

The speakers identify several areas where the Christian tradition offers what secular futurism lacks:

• A robust theology of personhood (imago Dei) grounding the distinction between humans, animals, and machines.
• An account of consciousness and intelligence that acknowledges their mystery rather than pretending to have solved it.
• An historical track record of warnings against utopian overreach.
• A solution to death (resurrection) that is not speculative but evidenced.
• A framework for meaning, morality, and human dignity that does not reduce to survival or utility.

The Call to Confidence

Lennox’s closing exhortation: Christians must not be paralyzed by fear of AI or intimidated into silence. The secular world is asking the right questions (What is a human being? What is intelligence? How do we govern transformative technology?) and Christians have serious, substantive answers. The moment demands theological literacy, public engagement, and the courage to “put heads above the parapet.”

Important Direct Quotes

“The artificial in artificial intelligence is real.” — Joseph McCrae Melly Champ (AI pioneer), as cited by Lennox

“We wouldn’t even know what it meant to produce a machine that is intelligent.” — Peter Norvig (AI textbook co-author), as cited by Lennox

“Everything made of words will be taken over by AI.” — Yuval Noah Harari, Davos Address

“The idea that AI are masters of words is false. They don’t understand a single word.” — Professor John Lennox

“He’s just getting away with saying what he believes without any serious intellectual or logical justification.” — Glenn (Host), on Harari

“Every attempt to build utopia avoiding the problem of human sin and rebellion against God has led to spectacular bloodshed.” — Professor John Lennox

Reference: https://www.youtube.com/watch?v=JYnuT83xk_g

Check out the Davos 2026 speech this article was based off of.