Tool or Terror? Artificial Intelligence and the Human Soul

Artificial intelligence has become an increasingly popular tool in recent years, with uses ranging from summarizing texts, to generating images, to taking notes for a virtual meeting—and, like any other significant cultural or technological innovation, this should prompt Christians to consider its implications carefully. While some suspicion of AIs like ChatGPT is already widespread because of practical concerns such as cheating on school assignments, it behooves us to also reflect on its potential spiritual effects: how using this technology may either benefit or harm our souls.

To begin with, AI is first and foremost a tool: something created by humans to serve a specific function (Merriam-Webster). It seems intuitive to us that many tools and technologies are neither intrinsically good nor evil, but rather dependent on how they are used—a can of paint may either be used for redecorating your house or for vandalizing someone else’s property, for example, but the paint itself has no intrinsic moral status. AI generally appears to function in the same way. Across the scope of various AI softwares, one can imagine both good and productive uses (writing an essay outline, generating new art as illustrations) as well as sinful ones (cheating on an essay, creating pornographic imagery)—and the difference seems to be almost entirely found in the choices of the user, rather than some quality of the program itself. When technologies bring about bad or evil results, broadly speaking, it is more often because they are furthering sinful inclinations of their fallen human users than because they are themselves bad, and AI seems to be no exception.

However, there is still cause for caution—AI, unlike many previous technologies, has the potential to mediate our lives in a way that can become idolatrous. AI is now accessible to us in an increasingly vast range of forms and functions, with entities like OpenAI (which developed ChatGPT) claiming that artificial intelligence “has the potential to benefit nearly every aspect of our lives” (OpenAI). Extending into even more areas than already pervasive technologies like the internet, AI can seemingly accomplish almost anything we want—recommending the next TV show we watch, deciding which social media posts we interact with, determining which stocks we should invest in (Guberti), or even driving our cars (Cummings). 

Because of this, we may be tempted to elevate AI in our minds to being something much more significant and universal than a mere tool, becoming dependent on it as a means through which we interact with the other areas of our lives. Instead of focusing ourselves on an attitude of reverence towards God (the way in which we should properly approach life), it therefore seems possible to slip instead into an attitude of dependence on technology. This phenomenon is probably a familiar one—anyone who has doomscrolled on Instagram or wasted hours on a YouTube rabbit hole has experienced something akin to what I am describing: allowing our everyday experience of life to involve more interaction with algorithms and AI (which like all tools, if unguided, can tend to primarily carry out our sin nature), than with prayer or scripture. In short, while there is nothing necessarily wrong with using AI for any morally legitimate use, we should also be very cautious to avoid using it as an all-encompassing “lens” through which we live our lives.

AI, across its many forms, is also uniquely generative, meaning that it can produce an incredible variety of seemingly novel content (essays, instructions, artwork, computer code). From a Christian standpoint, this is both good and bad. On the one hand, this property is good because it offers unprecedented potential for the furthering of human creativity and innovation—qualities with which God has endowed us, and by the use of which He is pleased. Indeed, in passages from the Bible like Exodus 25-27 (the building of the Tabernacle), God commands his people to engage in artistic expression for His glory—AI can fall solidly into this category as well, as a means by which Christians can exercise their skills and creativity to produce beautiful, God-glorifying media (ESV).

On the other hand, though, AI can easily produce the impression that it is doing something more  than human creativity is capable of—that we have fundamentally improved on our God-given ingenuity by creating this new technology. This is admittedly partly true—AI and algorithms often do serve computational functions that would be difficult (or impossible) for humans to achieve. However, it is also partly false. Although AI is generative, it is also derivative–that is, it is something created by humans. If (to borrow a term from J.R.R. Tolkien) God is the creator, and humans are “sub-creators” mimicking God in a limited way, then AI is a “sub-sub-creator,” drawing its content and inspiration from humans’ databases or webpages. In fact, AI is a step down in the creative hierarchy, rather than an advance upward (as many people seem to suppose)—it is absolutely a useful tool, but we ought not to expect it to either exceed the creative capacity or transcend the flaws of the already imperfect humans who created the material it draws from, any more than a human artist can improve on God’s work of Creation.

This leads to one final consideration: AI is malleable. This last quality is both one of AI technology’s most innovative aspects (it is designed to “learn” and improve over time, something that no other tool really does), and one of its most potentially harmful elements. Since the processes that compose any given AI are meant to adapt to human input (i.e. the desires of the user), this means that if the user is attempting to accomplish something sinful using the AI, the AI will probably increasingly facilitate that sin as it receives more inputs. (An immediate objection to this argument is the presence of limitations built into some AI programs, preventing negative applications such as disinformation, self-harm, or gambling. However, these restrictions are often easily circumvented, and with AIs’ increasing prevalence, many new, non-restricted AIs are also becoming available, allowing for practically any negative use imaginable (Chen; Nolan).) Regardless of whether individual AI softwares try to protect against perceived ill usage, then, it seems that AI taken as a whole is almost certainly adaptable enough to allow humans to accomplish almost anything they want—whether good or (in many cases) bad. Most tools have far narrower scopes of function, and as a result cannot make themselves more useful for sinful intentions; AI, if used wrongly, might well adapt and compound on human sin in a far greater way. If humans are naturally inclined to sin, then such a malleable tool ought to be treated cautiously: AI will shape its users just as much as its users shape it.

In conclusion, perhaps an accurate (if slightly cliche) summary of AI would be: “with great power comes great responsibility.” This essay is by no means intended to advocate strictly against using AI—such technology is an incredible outworking of human ingenuity, allowing for countless different, often beneficial uses. Nevertheless, a tool of such power and potential can be used to achieve extremely bad ends as well as especially good ones, and therefore merits careful consideration. As Christians, we are called to examine the spiritual consequences of all aspects of our lives in this way—and for AI especially, despite the current excitement over its applications, such thorough examination is certainly needed. 

Chen, Brian X. “Don’t Use A.I. To Cheat in School. It’s Better for Studying.” The New York Times, 30 June 2023, www.nytimes.com/2023/06/30/technology/ai-chatbot-study-aid.html.

“Definition of tool.” Merriam-Webster, www.merriam-webster.com/dictionary/tool#:~:text=%3A%20a%20handheld%20device%20that%20aids%20in%20accomplishing%20a%20task. Accessed 31 Jan. 2024.

Guberti, Marc. “Can AI Pick Stocks? A Look at AI Investing.” US News and World Report, US News and World Report, 16 Oct. 2023, money.usnews.com/investing/articles/can-ai-pick-stocks. Accessed 31 Jan. 2023.

Cummings, Mary. “What Self-Driving Cars Tell Us about AI Risks - IEEE Spectrum.” Spectrum.ieee.org, 30 July 2023, spectrum.ieee.org/self-driving-cars-2662494269.

Nolan, Beatrice. “There Are “Virtually Unlimited” Ways to Bypass Bard and ChatGPT’s Safety Rule, AI Researchers Say, and They’re Not Sure How to Fix It.” Business Insider, www.businessinsider.com/ai-experts-say-no-fix-jailbreaks-chatgpt-bard-safety-rules-2023-8. Accessed 31 Jan. 2024.

OpenAI. “Usage Policies.” Openai.com, 23 Mar. 2023, openai.com/policies/usage-policies.

‌Satariano, Adam, and Paul Mozur. “The People Onscreen Are Fake. The Disinformation Is Real.” The New York Times, 7 Feb. 2023, www.nytimes.com/2023/02/07/technology/artificial-intelligence-training-deepfake.html.

‌The Holy Bible: ESV, English Standard Version Containing the Old and New Testaments. New York, American Bible Society, 2007.

Tolkien, J.R.R. Tales from the Perilous Realm. HarperCollins, 2009.