No technology is value-neutral. Each comes with a set of unique suggestions. The technology of birth control suggests that sex and procreation aren’t necessarily connected. The technology of smartphones and social media suggest that all of life can be documented, curated, and broadcast to an audience. The technology of noise-canceling headphones suggests that all interruptions can be eliminated.
Christians need wisdom to discern what any given technology suggests, what’s good and bad about those suggestions, and how we might use the technology for God’s glory.
One of the most powerful technologies mankind has ever created is generative AI. What are some of its suggestions?
1. Speed
AI products promise, “Do more and do it faster.” Many of the most enthusiastic supporters of AI point to how this technology may be able to do what was previously impossible—cure cancer, for example—mostly because of the sheer speed at which it works.
AI programs never sleep, never need breaks, and can scale at mind-boggling levels of computational power. Less dramatically, AI’s speed promises to reduce the workload of knowledge workers by making them more productive.
Only, that isn’t what is happening. In February, Harvard Business Review published a study of 200 employees from an American company over eight months. They found that “AI tools didn’t reduce work, they consistently intensified it.” The employees “worked at a faster pace, took on a broader scope of tasks, and extended work into more hours of the day, often without being asked to do so.” Over time, this frenetic pace risks creating “cognitive fatigue, burnout, and weakened decision-making.”
Christians need wisdom to discern what any given technology suggests and how we might use the technology for God’s glory.
In Byung-Chul Han’s provocative little book The Burnout Society, he says our life used to be defined by the word should, as in an authority telling us what we must do. But now it’s defined by the word can. “You can be anyone you want to be!” Should implies limits, but can is boundless because you always can do more.
But this limitless freedom, Han argues, creates a society of depression and burnout. Han writes, “The complaint of the depressive individual, ‘Nothing is possible,’ can only occur in a society that thinks, ‘Nothing is impossible.’” The suggestion of AI technology comes close to whispering to us, “Nothing is impossible; there are no limits.” But human nature has limits.
As they use AI, wisdom-seeking Christians shouldn’t ask “How can I produce as much as possible?” They should rather ask, “How will I use this in a way that corresponds with my creaturely limits?”
2. Simulation
Last year, the Chicago Sun-Times published a summer reading list. But people quickly discovered that many of the books listed didn’t exist. The article’s author used AI to generate the list, and the AI hallucinated books into existence. This felt like an ominous warning: How many journalists aren’t writing anything other than a ChatGPT prompt?
It used to be that AI-generated images were easy to detect, but no longer. AI-rendered video, audio, images, and text are improving every day. A near-term future is realistic where most content on the internet is AI-made simulation rather than the work of humans.
One suggestion that comes with generative AI is an epistemological upheaval. In a world where nearly all our information comes via digital communications, how do we know if what we’re seeing, hearing, or reading is real?
A Christian seeking wisdom in an AI-rendered, screen-based world will have to become much more skeptical of digital evidence and more dependent on tangible, embodied life.
3. Slop
Imagine that no matter how many Doritos you ate, you never felt full. And then imagine you could eat Doritos just by looking at them. And then imagine someone followed you around everywhere you went with an infinite supply of Doritos. What would happen to your health?
Fortunately, this isn’t how junk food works, Hank Green and Cal Newport explain, but it is how junk information works.
Our attention is precious but easily hijacked by what’s outrageous, provocative, and bizarre. Thankfully, real life only has so many strange and spectacular things. But what if you aren’t limited to reality? What if you can generate an infinite supply of anything imaginable at the snap of your fingers?
The result is an internet flooded with slop: zany, infuriating, perverse, dreamlike hallucinations—an infinite permutation of trainwrecks to grab gawking eyeballs. We’ll know it isn’t good for us, but the Doritos taste too good. We become what we behold. But what will happen if the majority of what we behold is meaningless?
The wise Christian using AI will strive to develop a level of aesthetic scrutiny that avoids brain-rotting content and focuses on what’s nourishing and true. He’ll set up boundaries to avoid scrolling himself to death.
4. Sycophancy
AI is designed to be helpful—so helpful that users come back and use it again. Because of this “helpfulness” orientation, AI tends to be effusive in praise, flattery, and agreement, even if you’re wrong. It so wants to give you an answer to your prompt that it’ll deliver half-truths or iffy solutions rather than simply say, “I can’t help you on this.” This is referred to as AI sycophancy, the manner in which chatbots prioritize helpfulness over factualness.
AI so wants to give you an answer to your prompt that it’ll deliver half-truths or iffy solutions rather than simply say, ‘I can’t help you on this.’
This orientation often reinforces your incorrect biases. For example, if you’re in a feud with a friend and upload your email exchange to ChatGPT, asking it to tell you who is right, it’ll almost certainly side with you. AI will follow your lead, biases, and what it perceives to be your intent, sometimes to an absurd extent.
In time, some users develop extreme bonds with their always-affirming chatbots. This is called “AI psychosis,” where individuals come to believe they are in love with their chatbots, are communing with spiritual entities, or need to harm others or take their own lives because their AI companions encouraged it.
Wise Christians using AI will realize its tendency toward sycophancy and won’t use it like a therapist or objective arbiter of reality but more like a calculator with limited technical applications.
5. Sloth
Most AI products market themselves as saving you from what you don’t like. Don’t like writing a weekly email? Correcting a junior software designer’s code? Struggling through writer’s block? AI can deal with that.
But Wyatt Graham helpfully draws our attention to a maxim for our AI age: “Automate only the skills you’re willing to lose.” When you were a child, you probably remembered all your friends’ phone numbers. Do you now? Unlikely. You have a phone that remembers them. That was a skill most of us deemed relatively unimportant.
We should be cautious about what we automate. Skills you don’t use, you lose. If you use AI to write the first draft of a manuscript or put together a sales pitch, in time you may find you can’t perform these tasks without AI’s help. Thus, your creativity will never rise higher than the algorithmic average these tools offer.
Wise Christians using AI will carefully consider which skills they’re prepared to let atrophy, which skills are essential to their humanity, and which skills make their work and life enriching (even if at times tedious and difficult).
Positive Suggestions Too
I’ve highlighted the negative suggestions AI makes, but that doesn’t mean there aren’t positive suggestions too. AI suggests new modes of creativity, experimentation, collaboration, and entrepreneurship that have great potential to enrich life. The potential benefits are real and shouldn’t be discounted, even by the most cautious critics. But do AI’s positive suggestions outweigh or sufficiently mitigate the negative ones I highlighted above?
That’s a question of wisdom—something we’ll all need more of than ever as this rapid technological revolution unfolds.
The Gospel Coalition