Why AI Is Our Latest Golden Calf

The form of human idolatry changes. The propensity for idolatry does not.

Whether building sky-high towers that reach God, as in the tale of Babel in Genesis 11; making golden calves in Exodus 32; or raising Nebuchadnezzar’s golden image in Daniel 3, humans tend to project God-like power onto human-made artifacts. The great theologian John Calvin likened the human heart to a “perpetual factory of idols.”

Enter the latest chapter of this deeply human tendency: AI. This is the next widget on the conveyor belt of idols. What’s fascinating is that even AI experts admit they aren’t even sure exactly how AI works. If they can’t, I can’t either. From all the available summaries I’ve read, AI works by algorithmic prediction based on the enormous quantities of data it has scraped.

Questions now abound about whether, and to what degree, AI possesses God-like attributes — such as aseity (existence), infallibility (without error), and omniscience (all-knowing) — and God-authored human capacities, such as rationality, moral responsibility, consciousness, sentience, and ensoulment.

Now that conversation is accelerating, as the New York Times reports in a stunning story about Anthropic consulting with religious leaders on the moral status and religious implications of AI agents. The story shows that the secularized tech vanguards of Silicon Valley — who largely assume a materialist account of reality — are courting and pressing religious leaders to consider the outer boundaries of their own assumptions about AI’s moral status as a conscious being. Other stories give eerie indication of where AI accelerationism is going: “If we build a superintelligence, we will first fall in love with it. Then we will worship it. Finally, we will make sacrifices to it.” Thankfully, not all of the AI champions are singing from the same technocratic hymnal. Sam Altman, OpenAI’s lead, is right to point out the errors of placing religious burdens on AI: “I am very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models, and think it is a real safety issue.” Restraint and sobriety in the face of unbounded optimism is a welcome chastening.

As of this writing, I am hard-pressed to find many — or any — serious religious leaders entertaining AI as a morally responsible person or an agent that can channel the divine. Earlier in the year, the Vatican released a sprawling encyclical insisting upon the unique status of humanity, a unique status that no man-made machine can fully replicate. Far ahead of our current craze, I was part of a group of Christian scholars and leaders who, in 2019, released a statement denying “that any part of creation, including any form of technology, should ever be used to usurp or subvert the dominion and stewardship which has been entrusted solely to humanity by God; nor should technology be assigned a level of human identity, worth, dignity, or moral agency.” What seemed like far-off speculation a mere seven years ago now seems the topic du jour.

Our argument was and remains: AI is not conscious. It lacks self-awareness or the ability to conceptualize self-transcendence. AI does not possess moral free choice, that is, the ability to deliberate and make choices all on its own. AI is not “alive” in any sense that bears resemblance to organic beings. AI is not sentient; it cannot feel. AI has neither soul nor body. It has no reflexive rationality that was not first programmed.

Disembodied, it lacks the foundational requirement of what makes us human persons and created in the divine image — the Body-Soul unity. Cognitive simulation without a human body does not constitute personhood. Biblically, human personhood is the unity of body and soul.

The fact that AI must be trained for “alignment” (industry lingo for setting proper moral parameters) means they are not innately morally responsive. Lacking moral agency — needing training to know right from wrong — highlights a salient difference between humans and machines: moral awareness is innate in humans (Rom. 2:14-15).

As philosopher Christopher Tollefsen helpfully summarizes, AI may have the properties of mind for information processing and still lack consciousness. From this, we can conclude what AI does: it mimics cognition. Simulating cognition is different from possessing cognition in itself. It has the “illusion” of consciousness, as Tollefsen states, that allows predictions to follow from inputs. That may be as much as we can say.

Apart from the powers of computation that humans give to AI when they create it, AI would never do. Because it has no property of doing apart from the humans that empower it, it has no property of ontological being, either. There are also real human costs to ascribing human-like powers to AI: We begin to assign worth based on function. Here, as the cognitively disabled, the comatose, or those with dementia are edged out for what qualifies as “humanity,” the logic of anti-humanism will cash out in ascribing human agency to machines at the cost of safeguarding our own dignity.

AI cannot innately reason about universals, freely choose, or have intent — three qualities Tollefsen requires for consciousness to exist. And what I mean is that AI may appear to do those things, but the appearance is merely materially created. This is a decisive point to linger on because it shows that the comparison between humans and machines is not an apples-to-apples comparison of reasoning ability, but an apples-to-oranges comparison that speaks to the differences in being.

Consider how a human comes into being compared with a machine. While persons create a human embryo, it possesses all the qualities necessary to develop from its own internal nature; AI does not. AI has no internal “propulsion” to direct itself apart from a human controller. Unlike humanity, which is self-directing and has its own natural ends, machines only direct insofar as we impose our control on them.

As Tollefsen writes, “How can what is not conscious — namely, physical reality — generate something that is metaphysically different from itself — namely, the phenomenal properties of consciousness? This would appear to violate a fundamental principle of reality, that one cannot get something from nothing.” AI would not exist without human engineering ability.

The conversation carries further irony. The AI materialists who often deny the existence of the soul — and whose materialism cannot adequately explain consciousness, selfhood, or even a reliable perception of reality—now want to attribute soul-like capacities to machines. This is worldview chaos.

In Christian theology, the incarnation is decisive in this conversation. Christ became flesh, not code, tokens, or compute. Christ was fully God and fully man. Christian theology places a premium on creation. Matter matters. Outsourcing humanity to machines and uploading it to machines traffics in the same type of gnostic dualism that separates body from the Self, the same thought paradigm that licenses transgenderism, for example. If we are not our bodies, we’ve lit the spark for equating personhood with non-bodily reality.

Psalm 115:4-8 speaks to the current cultural phenomenon we find ourselves in. There, the Psalmist mocks the impotence of man-made objects that we offload our hopes and dreams onto:

[4] Their idols are silver and gold,

the work of human hands.

[5] They have mouths, but do not speak;

eyes, but do not see.

[6] They have ears, but do not hear;

noses, but do not smell.

[7] They have hands, but do not feel;

feet, but do not walk;

and they do not make a sound in their throat.

[8] Those who make them become like them;

so do all who trust in them. (ESV)

Notice that the Psalmist mocks nations and cultures that forge idols that resemble human nature’s capacity but lack their fulfillment. The lesson of idolatry in Scripture is that people remain in perpetual disappointment when they try to equate man-made artifice with excellence and ability. When man forgets God, he does not forget his ability to worship — he merely redirects it to that which can never truly satisfy.

It is now common to hear the moral call not to offload one’s critical thinking to AI in the name of “cognitive surrender.” That’s true, but the same truth applies to not offloading and surrendering human agency or divine agency to AI.

***

Andrew T. Walker, Ph.D., is Dean of the School of Theology at The Southern Baptist Theological Seminary and Professor of Christian Ethics and Public Theology. He is a Fellow with The Ethics and Public Policy Center and the Managing Editor of WORLD Opinions. You can follow him on X: @AndrewTWalker.