There is much in Jonathan Barlow’s piece that resonates with me. I appreciated how he cites some seminal thinkers like Neil Postman, Jacques Ellul, E.F. Schumacher, and Ivan Illich as he frames his argument. In particular, I think Barlow is insightful in identifying the pitfall of “commensurability” when comparing humans and AI in terms of performance. I am reminded of the words of the famous computer scientist Edsgar Dijkstra, who suggested that the question as to whether machines can think is about as relevant as the question of whether “submarines can swim.”[1] Barlow is absolutely correct in suggesting that “looking for human exceptionalism in the excellence of our output relative to machines” is an “erroneous concession.” We are, as Barlow emphasizes, created in the image of God (and here I was delighted to see him cite one of my favorite theologians, Herman Bavinck). Consequently, Barlow emphasizes the importance of a Christian anthropology as a starting point, which I believe is crucial. In my own writings on this topic, I have argued that “it is the ontological questions that will help us discern what separates humans from machines. We are often captivated by what things can do, rather than asking what things are.”[2] In addition, I appreciate Barlow’s warning against idolatry, which is a real pitfall when dealing with powerful technologies.

However, there are a few basic statements that Barlow makes with which I would quibble.

First, I am not sure that framing AI as “imago Hominus” is helpful. Barlow describes AI as “crafted in the image of man” in at least three ways. Later he contends that “Christians are unique in being able to explore unanxiously the full potential of AI systems as social things that are in our image and able to accompany us in our mission.” The notion of creating something in our own image gives me pause; it has echoes of a god-like pursuit, a theme explored in Mary Shelley’s Frankenstein, where the creature famously declares “Remember, that I am thy creature; I ought to be thy Adam…”[3] While I don’t dispute that this impulse may be motivating many who are creating AI, I don’t think this is a helpful frame for a Christian perspective; it will lead to blurring the lines between humans and machines.

Rather than imaging humans, I like Marshall McLuhan’s more limited notion that our machines extend or amplify some limited human capacities.[4] For example, writing extends the memory, the car extends the foot, the telescope amplifies the range of the eye, etc. Likewise, Large Language Models amplify our ability to gather and sift through information, and machine learning extends our ability to classify statistical patterns. While an AI chatbot may convincingly mimic a human, it extends a very limited aspect of our humanity. Moreover, McLuhan warned that while our tools amplify certain skills, those same skills atrophy in the person using the tool (a condition he calls “autoamputation”).[5] McLuhan warns that writing does not improve memory; it is memory outsourced and diminished. The car extends the foot, but driving makes us less fit to walk. The implication for AI is that we need to be on guard against the atrophy of our cognitive abilities from cognitive offloading. Barlow acknowledges this same fact when he writes that “humans tend to offload judgment and conscience until we become an atrophied node of our own artifact.”

In the section on AI “crafted in the image of man,” Barlow discusses how AI can undergo a kind of “catechesis through fine-tuning methods to reinforce courtesy and helpfulness.” This is achieved through a process called “reinforcement learning” to enhance certain characteristics in chatbots, which can be helpful. Even so, I have argued that it is important to emphasize that, unlike a human, an AI can never be virtuous even though there can be a kind of “virtue by proxy” in which AI can be “trained to mimic certain virtues.”[6]

At one point, Barlow insightfully poses the question of how to uphold the “maker-AI distinction” and the “otherness” of an AI system. He asks whether we need to avoid personal pronouns for machines or make other design choices. While Barlow poses these as open-ended questions, as a computer science professor I would argue fervently that we need to design machines in a way that keeps this distinction: people are people and machines are machines. I tell my computer science students not to create machines that pretend to be human.[7] Elsewhere I have written that by creating machines that mimic people, we risk something I like to call “ontological confusion.”[8] In posing these questions, Barlow is identifying a crucial area for future research: namely determining appropriate design norms for responsible AI.[9]

Finally, near the conclusion, Barlow writes “…the New Jerusalem Christians are building is a city…” I think it is important to note that we do not build the New Jerusalem at all—it is a city that comes “down out of heaven from God” (Rev. 21:2) and “whose architect and builder is God” (Heb. 11:10). A Christian eschatology needs to sharply distinguish itself from the ideologies of many AI leaders who think we can harness AI to usher in a kind of Utopia (see, for instance, Marc Andreessen’s article on “Why AI Will Save the World”). Instead, I think we are called to be a faithful presence by “making some imperfect models of the perfect world to come.”[10] We are signposts of a new creation, but the actual construction is not subcontracted to us. Even so, some of the honor and glory of the nations, free from sin, may find its way into the New Jerusalem as suggested in Revelation 21:26.[11] And, as I’ve suggested elsewhere, that may even include AI.[12]

Despite some of these quibbles with Barlow’s wording, I think Barlow successfully shows that the Christian faith has some crucial insights for the world of AI. I echo Barlow’s call to be alert to the issue of incommensurability and to receive the gift of AI with “thanksgiving rather than fear.”

Derek Schuurman is Professor and Department Chair of Computer Science and Faith & Technology at Calvin University.

NOTES

[1] Edsger W. Dijkstra, “The Threats to Computing Science” (technical report, EWD898, Technical University Eindhoven, Eindhoven, Holland, 2003).

[2] Derek C. Schuurman, “Artificial Intelligence: Discerning a Christian Response,” Perspectives on Science and Christian Faith 71, no. 2 (2019): 79. Emphasis in original.

[3] Mary Shelley, Frankenstein, Dover Thrift Editions (Dover Publications, 1994).

[4] Marshall McLuhan, Understanding Media: The Extensions of Man, 10. print, Media Sociology (MIT-Press, 2002), 45.

[5] McLuhan, Understanding Media, 44.

[6] Derek C. Schuurman, “Virtue and Artificial Intelligence,” Perspectives on Science and Christian Faith 75, no. 3 (2023): 155, https://doi.org/10.56315/PSCF12-23Schuurman.

[7] Schuurman, “The Problem with Chatbot Personas.”

[8] Derek C. Schuurman, “The Problem with Chatbot Personas,” Christian Scholar’s Review Blog, November 4, 2024, https://christianscholars.com/the-problem-with-chatbot-personas/.

[9] I explore some design norms for AI in “The Challenges and Opportunities of Artificial Intelligence: Reflections from a Christian Computer Scientist,” Christian Higher Education 25, nos. 1–2 (2026): 92–106.

[10] Lewis B. Smedes, My God and I: A Spiritual Memoir (W.B. Eerdmans, 2003), 59.

[11] For a further discussion of this, see Richard J. Mouw, When the Kings Come Marching in: Isaiah and the New Jerusalem, Rev. ed (W.B. Eerdmans, 2002).

[12] Derek C. Schuurman, “AI Won’t Get Us to Heaven. But It Might Be There When We Arrive.,” Christianity Today, October 21, 2025, https://www.christianitytoday.com/2025/10/ai-in-heaven-technicism-dataism-new-creation/.