Back in February, I posted a few thoughts about artificial intelligence on Facebook. I found myself somewhere between the optimism of Star Trek and the anxiety of The Terminator.
I ended with a little apocalyptic humor: “No matter how smart it gets, we’d better be the ones holding the kill switch.” (Feb. 2 FB post)
The joke carried a serious concern. We are developing tools of extraordinary capability, and we need to retain responsibility for what we build and how we use it. A few months later, I am still thinking about that responsibility.
I understand the reluctance surrounding AI. Questions about jobs, creative ownership, privacy, infrastructure, and human control deserve a hearing. So do the concerns of parents and teachers trying to understand what these tools will do to children who are still learning to think.
I also understand the excitement. I have experienced some of it myself.
What troubles me is the possibility that people with serious moral convictions will conclude that their only faithful response is to leave the conversation. If good men and women surrender their place in the discussion, who will speak for the people whose dignity and well-being ought to govern it?
We need to remain engaged. And our engagement needs a purpose deeper than keeping up.
What stewardship requires
I have often said that technology is essentially neutral—raw potential waiting for human intention. I still believe that human intention matters enormously. A tool can help us serve a neighbor or exploit one.
But that statement needs qualification. Technology comes to us with choices already built into it. Let's not be naive about that..... Someone has decided what information it collects, what behavior it encourages, what success looks like, and who retains control. Responsibility begins before the user ever touches the keyboard.
Stewardship reaches across that whole process. It asks what we should build, how we should build it, who should benefit, and what limits we should accept.
It also requires us to hear objections fairly. A community questioning a data center may be asking reasonable questions about its resources and infrastructure. A musician concerned about the use of creative work may be defending something valuable. A teacher questioning an AI assignment may be protecting the learning that assignment was supposed to produce.
The International Energy Agency projects that global data-center electricity consumption will roughly double between 2025 and 2030. These tools have a physical footprint, and their costs belong in the conversation alongside their possibilities. www.iea.org
Calling every objection “fear of progress” would be a failure of stewardship.
So would using the word “inevitable” to end the discussion. AI may continue to develop, but that does not make every application, business practice, or classroom use inevitable. Human choices still matter. We must insist that they do.
History gives us questions
We have encountered disruptive changes before.
The printing press helped carry the Protestant Reformation through pamphlets and other printed works that reached audiences far beyond a single pulpit. It also carried competing religious campaigns and polemics. Expanding the reach of words did not settle which words deserved belief.
Industrial machinery presents another complicated example. The Luddites are often remembered as people who simply hated technology. They were skilled workers protesting machinery used to replace their labor and drive down wages. Their grievances deserve more respect than the label usually receives.
A development can increase production while imposing severe costs on particular people. We ought to remember those people when we speak enthusiastically about progress.
There are also examples of participation that included deliberate restraint. In the 1970s, researchers working with recombinant DNA confronted risks arising from their own discoveries. A voluntary pause in certain experiments, the 1975 Asilomar conference, and subsequent NIH guidelines helped establish conditions under which the work could proceed. (see note at the end about this conference)
That example matters to me. People involved in developing a technology accepted responsibility for questioning and limiting it.
History cannot promise that AI will turn out well. It can help us recognize that curiosity, criticism, caution, and responsible development all have a place in the discussion. Staying involved sometimes means insisting that a particular project slow down—or stop.
The Princeton Response
There is a related lesson in the Christian response to Darwin.
When On the Origin of Species appeared in 1859, Christian responses varied. Some opposed it, some sought ways to accommodate its findings, and some supported it. The familiar picture of a single confrontation between “science” and “the church” leaves out a much more complicated conversation.
Charles Hodge, the Princeton theologian, published What Is Darwinism? in 1874. He examined the arguments and distinguished evolution and natural selection from what he understood as the exclusion of divine design. That last issue stood at the center of his objection. I remember the 1st time I read Hodge's response- it was full of retrained respect and logical analysis- we miss that so much in our drama filled bantering today!
Asa Gray, a devout Presbyterian and botanist, reached a different judgment. He engaged extensively with Darwin and argued that natural selection need not exclude divine purpose.
I am not trying to reopen that entire debate here. What interests me is the willingness to understand the argument, identify the assumptions, and bring Christian conviction into a serious exchange.
Hodge and Gray did not agree. Both participated.
That is a useful reminder for us. Faithful engagement will not produce identical conclusions. It should produce better questions, more careful distinctions, and disagreements that address what someone actually believes.
We need that discipline when discussing AI. A demonstration of what a system can accomplish is one thing. A claim about what that accomplishment means for human identity, moral responsibility, or the future of civilization requires further examination.
We should understand the capabilities and examine the promises being attached to them.
Babel and the Garden
The biblical framework that keeps returning to me is the relationship between Babel and the creation mandate.
In Genesis 1, humanity receives the command to fill the earth and subdue it. Genesis 2 places the man in the garden to work it and keep it. Human beings have work to do in God’s world: cultivating, discovering, ordering, and caring for what has been entrusted to them.
That responsibility gives me reason to welcome discovery. It also places discovery under an authority greater than our appetite for achievement.
Genesis 11 gives us another picture. The builders of Babel are organized, cooperative, and technically capable. They know how to make bricks and construct something impressive. Their stated ambitions include making a name for themselves and avoiding being scattered.
Read alongside the command to fill the earth, their project raises questions about self-exaltation, control, and resistance to God.
Those questions reach into our own ambitions. We can build something remarkable and still build in the wrong direction.
The creation mandate does not give blanket approval to everything we can invent. It gives us responsibility before God for what we do with His creation.
That is why purpose and limits matter. Does this use of AI help people flourish? Does it serve the vulnerable? Does it preserve accountability? Are we cultivating what has been entrusted to us, or seeking a kind of security and significance that leaves no room for dependence on God?
Those questions apply to the companies developing AI. They apply to me when I use it.
My own experiment
I cannot discuss this subject as a detached observer. I have used AI in making music, improved research into many topics, editing and discussion of ideas and I have enjoyed the process of bringing lyrics and musical ideas into something people can hear. If anything, I do it for me- healthy messages make a more healthy disposition. And I can't find new music that has a lot of good messaging.
That enjoyment gives me a stake in the argument. I need to acknowledge it. I should be willing to examine a practice even when it helps me do something I love.
The responses have been interesting. So far, I have experienced a warmer reception on SoundCloud than on Spotify. I have wondered whether differences in the audiences contribute to that response, but it is too early for me to draw a conclusion. Discovery, genre, audience size, and many other factors could be involved. And I have to affirm that I understand REAL musicians distaste for it- but I do firmly believe I am creating things that would not exist without my initiation and work- but I digress....
What I can say is that the experience has made the questions personal.
What is the listener seeking? Some people are moved primarily by what they hear. Others place considerable value on knowing that a person physically performed the music. Both deserve honesty about the process.
I need to respect other creators, describe my own contribution truthfully, and remain open to criticism. Making something meaningful to me does not settle every question about the tools that helped me make it.
That willingness to examine my own enthusiasm is part of the stewardship I am asking others to practice.
How do we know? My drum beat for an updated "Epistemology"
As an educator, I think one of our most urgent responsibilities is to teach students how to judge what deserves their trust.
I have written about epistemology—the study of knowledge and justified belief—for years. I have also found it difficult to generate much interest in the subject. (here is the start of a series here- Epistemology series intro )
Perhaps the word sounds too academic. The problem is immediate.
A student encounters an answer that is polished, confident, and apparently well supported. What should the student do next?
AI gives fresh urgency to questions education has always needed to address. What is being claimed? Where did the information originate? Does the evidence support the conclusion? What is fact, what is interpretation, and what remains uncertain?
In my earlier writing, I called for these habits to be woven throughout education, along with metacognition: examining our own thinking. We need to investigate the information in front of us and the reasons we find it convincing. www.jayopsis.com
Am I accepting this because the evidence is strong? Does it agree with something I already want to believe? Have I mistaken confidence for competence? What would lead me to change my mind?
An updated application of epistemology should make those questions practical.
In history, students can trace a claim to its source and consider its context. In science, they can distinguish observations from proposed explanations. In literature, they can defend an interpretation with evidence from the text. In a Christian school, they can learn to examine theological claims carefully and read Scripture in context.
These habits require subject knowledge. Students need enough understanding to recognize when an answer is questionable. Reading, writing, calculation, memory, and sustained attention give them resources with which to assess what a tool produces.
We should protect the effort that develops those abilities.
A student might use AI to explore a question and then verify its claims, identify omissions, and defend a revised conclusion. The teacher can assess the reasoning and correction alongside the finished answer.
The educational goal is a person capable of responsible judgment.
Teaching students to operate a powerful tool while neglecting that judgment would leave some of our most important work unfinished.
Remaining responsible
My Facebook joke about the kill switch still expresses something I believe: human beings must retain meaningful authority over the systems they create.
But I also recognize the limitation of the image. A complex technology spread across many institutions will not necessarily have one accessible plug. We need responsibility built into its development and use long before an emergency arrives.
Someone must remain accountable for a decision. Someone must be able to hear an appeal, investigate a failure, correct an error, and stop a harmful practice. “The AI recommended it” cannot be the end of the explanation.
There must also be room for conscientious refusal. A person who declines a particular tool may have reasons we need to hear. Staying in the fray can include principled opposition. I am asking us to remain responsible participants in the conversation, willing to understand and willing to challenge.
I do not know how this technological transition will unfold. Its pace makes prediction difficult, and uncertainty should make us humble.
I do know that greater capability brings greater responsibility.
We need good men and women who will learn enough to ask informed questions, speak for those bearing the costs, protect the development of human judgment, and insist on limits when limits are needed.
As Christians, we can approach that work with hope rooted in God, with an honest understanding of human fallenness, and with care for the neighbors our choices affect.
We have a garden to tend. We also have a tendency to build towers for our own glory.
May God give us the wisdom to examine what we are building, the courage to remain involved, and the humility to receive correction along the way.
End Note: The 1975 Asilomar conference leaves a question that technical safeguards cannot answer by themselves: Who should help decide what risks are acceptable, and who bears those risks?
For AI, that would mean bringing educators, artists, workers, families, and other affected communities into the conversation alongside developers. A laboratory containment plan has a relatively concrete purpose; deciding how AI should influence learning, livelihoods, creativity, and trust requires a broader moral conversation.










