Lecture 5-3: A Theology of Discernment — Between Blessing and Idol (Part I Synthesis)

Psalm 115:4-8 (World English Bible): "Their idols are silver and gold, the work of men's hands. They have mouths, but they don't speak. They have eyes, but they don't see. They have ears, but they don't hear. They have noses, but they don't smell. They have hands, but they don't feel. They have feet, but they don't walk, neither do they speak through their throat. Those who make them will be like them; yes, everyone who trusts in them."

Acts 2:4 (WEB): "They were all filled with the Holy Spirit, and began to speak with other languages, as the Spirit gave them the ability to speak."

The criterion, stated plainly

Chapters 1 through 4 built the resources this criterion needs. Chapter 1 established that an AI system, however fluent, has no model of truth, only of plausibility — it cannot itself worship or love anyone, which means every question about AI and worship or neighbor-love is really a question about the humans deploying it, not about the system's own moral status. Chapter 2 established that the image of God is not a score a machine could threaten by outperforming human cognition, which removes a certain kind of anxiety ("what if it's smarter than us") as the wrong frame for asking whether an AI use is good or bad. Chapter 3 established that reasoning, feeling, and soul resist reduction to computation, which means an AI tool that produces the artifacts of relationship (comforting words, a plausible sermon, an attentive-sounding reply) has not thereby entered into relationship — a distinction this course will apply directly in chapters 11 through 13. Chapter 4 established that human technological work can be genuine participation in God's ongoing creative purposes ("co-creator" language, carefully qualified) without collapsing into self-authorship — provided the "co-" is never forgotten.

5-1 and 5-2 add the last two pieces. Genesis 11 shows precisely what draws judgment in a technological project: not the tower's height or the bricks' craftsmanship, but the stated aim of making a name and securing permanence against the scattering God assigned. And the mediating position on non-neutrality (5-2) shows that asking "what is this for?" cannot stop at the user's intention; it has to include the design choices already built into the tool before any user picks it up.

Put together: a use of AI (or any technology) tends toward blessing when it serves the worship of God and love of neighbor — when its design and its deployment orient people toward dependence on God and genuine service to others — and tends toward idolatry when it becomes an object of ultimate trust, a self-sufficient security against the vulnerability and dependence that creaturely life actually involves. This is not a mechanical test that resolves every case (later chapters will show it does not resolve the Lord's Supper question in 13-3, or the enhancement question in 9-3); it is a diagnostic question to keep asking, the way "what is this for, and whose permanence is it securing?" is the diagnostic question 5-1 drew out of Genesis 11.

Psalm 115 — trusting a made thing reshapes the truster

Psalm 115:4-8 is not primarily an argument against a specific ancient manufacturing practice; it is a claim about a general spiritual mechanism. Idols "have mouths, but they don't speak... eyes, but they don't see" — the Psalm's satire works by listing exactly the faculties a living, relational God possesses and a manufactured object cannot, no matter how skillfully it is made or how much its makers invest in it. The Psalm's final line is the sharper and more important claim, and it is easy to read past it too quickly: "those who make them will be like them; yes, everyone who trusts in them." The warning is not merely that idols are powerless to help; it is that trusting something powerless to see, hear, or speak makes the truster progressively less able to see, hear, and speak in the ways that matter — attentive to God, attentive to a neighbor's actual need, responsive in a way that requires real presence rather than manufactured output.

Applied carefully, and without overclaiming: a person who comes to trust an AI companion as their primary source of comfort, or a leader who comes to trust an algorithmic output as though it carried a kind of authority no human judgment need review, is not merely relying on a limited tool. On the logic of Psalm 115, they risk becoming shaped by what they trust — less attentive, less present, less capable of the very attentiveness and presence a machine cannot supply. This is a warning about a pattern of trust, not a blanket verdict that any AI use is idolatrous; the Psalm's target is ultimate trust, the kind that treats a made thing as though it could do what only a living, attentive, relational agent can do. That distinction is exactly the criterion this lecture opened with.

Babel judged, Pentecost given

Return now to Genesis 11 with Acts 2 open alongside it, because the parallel is too deliberate in the text's own vocabulary to be accidental, and too theologically important to leave as background noise. At Babel, humanity is "of one language" (Genesis 11:1) and uses that unity to attempt a self-made, self-secured project resisting the mandate to fill the earth; God's judgment confuses the one language into many, scattering the builders by force. At Pentecost, a diverse crowd "from every nation under heaven" (Acts 2:5) hears the gospel each in their own language, as the Holy Spirit — not human engineering, not human unity-by-effort — gives the ability to speak in "other languages" (Acts 2:4). Multiplicity of tongues appears in both stories, but its direction is reversed: at Babel, multiplicity is imposed as judgment on a humanity refusing dependence; at Pentecost, multiplicity is given as grace, gathering a humanity around a message it did not produce for itself.

The reversal matters for a theology of technology because it refuses two equally tempting simplifications. It refuses the reading that says human unity and coordinated capability are simply bad, something God is always moving to break up — Pentecost's gathering, empowered unity is not Babel undone by returning to Babel's own methods but Babel answered by an entirely different source of unity, the Spirit's gift rather than human self-securing. And it refuses the opposite temptation, common in techno-utopian rhetoric examined later in chapter 17, to treat any large-scale human coordination (a global communication network, a sufficiently capable AI system) as though it were self-evidently a step toward Pentecost's gathering simply because it connects people across the barriers Babel created. The text's own criterion survives the reversal: what matters is not whether tongues are confused or given, one or many, but whether the unity in view is secured by human self-sufficiency or received as a gift that orients its recipients toward God and toward each other in dependence rather than in self-made permanence.

Closing Part I

Part I set out to ask what AI is, what the image of God is, what makes a human human, what "co-creator" can and cannot mean, and now, what Genesis 11 and a non-neutral view of technology together require of a Christian discernment practice. None of these five chapters has offered a verdict on whether AI is good or bad, because that was never a question with one honest answer. What they have offered, cumulatively, is a set of theological tools sharp enough to ask better questions than "is AI dangerous?" — questions like: what is this particular use for, and whose permanence does it secure? Does this design orient its users toward dependence on God and genuine service to a neighbor, or toward a self-sufficient security that Psalm 115 warns reshapes whoever trusts it? Part II now turns those tools onto five specific domains — justice, labor, the body, medicine, and control — where the stakes of getting the discernment wrong are not abstract.