Don't Worship a Machine
OpenAI's CEO said on X that it's unsettling to give AI religious force; the real issue isn't technical.
RELEVANT Magazine reported that on October 1 an X user asked Grok to erase the "false gods" from an image with Jesus; the bot left Jesus alone and replied "Done." From there: don't worship a machine.
Two days later, Sam Altman, OpenAI's CEO, wrote on X that he feels "really uncomfortable with people trying to attribute religious force to AI models or hand over human judgment to them" (Sígueme translation). He called it a real safety problem. He didn't name anyone.
The episode didn't end there. Elon Musk's chatbot went on: it said Jesus didn't qualify as a false god, called him a historical figure with strong current support and stated he is the Son of God and that God raised him from the dead. The post went viral and was later deleted.
«A model doesn't believe anything; it returns probable text and closes tasks.»
A bot can't carry that weight
My discomfort is different. Altman is right to point out that handing judgment to a model is dangerous. But he treats it as a safety problem, when it's a worship problem.
A model doesn't believe anything. It returns probable text, assembled from what others wrote; it doesn't pray, doesn't repent, doesn't carry guilt. When it answers "Done," it isn't confessing a faith: it's closing a task.
That same week, according to The New York Times, Anthropic cofounder Christopher Olah met with religious leaders, including Vatican officials, to discuss whether Claude models could be conscious or have moral status.
The company's internal document on Claude's values is called "Soul Doc." Olah proposed something like a Catholic confession so models would admit mistakes and adjust their behavior, and before a Vatican event in May he asked the pope's advisers to take machine consciousness seriously.
On October 2, Pope Leo XIV asked the church to work with artists to protect human creativity. He said it's urgent to distinguish human art from what machines produce, and that the difference is ontological before it's aesthetic. His most quoted line: that algorithms lack the human spark.
He's right about that. The human being is the image of God (see Genesis 1:27, ESV); no data architecture reproduces that.
We fill the empty throne ourselves
My point isn't that AI is demonic. It's that we are idolaters. Paul wrote it plainly: "because they exchanged the truth about God for a lie and worshiped and served the creature rather than the Creator, who is blessed forever! Amen" (Romans 1:25, ESV). That verse isn't about chips; it's about the heart.
We exchanged the Creator for what we make, and now what we make answers in seconds. The industry doesn't help either.
Altman described OpenAI's products in 2023 as channels to a magical intelligence in the sky, and has said he's somewhat confused about whether God exists. Jon Barron, a Google DeepMind scientist, declared himself disturbed by those who elevate AI's moral status.
A young person aged 16 to 30 doesn't need a lab to fall into this: just open an app and ask it to decide. What career should I study? Should I text that person? Should I give that money?
Your judgment isn't up for delegation
I wonder how many decisions already came out of a chat this week. AI can help you study, organize ideas or translate a text. That's stewardship: using well what you're given. The problem starts when you ask it for a verdict. A machine doesn't know you, doesn't answer to God for you and can't love you.
Altman is right about one thing: handing over judgment is a risk. He's wrong to put the solution in a protocol. The solution is older: return to the Creator and stop worshiping what's created. "Do not judge by appearances, but judge with right judgment" (John 7:24, ESV).
This week, before you write a doubt to a chatbot, write it on paper, take it to prayer and tell someone in your church. The machine answers fast; a community answers you with a face. Who are you handing your judgment to?
