Navigating Artificial Intelligence with Biblical Discernment
Slides can be found here.
Links that Tim mentions about his AI tools are:
biblehunter.org
learntheword.net
It seems like just yesterday that Artificial Intelligence (AI) was a concept confined to the realm of science fiction. Yet, almost overnight, it has woven itself into the fabric of our everyday lives. From auto-generating routine work emails and organizing grocery lists to drafting complex computer code and analyzing complex datasets, AI tools have rapidly transitioned from futuristic novelties to everyday essentials.
For many Christians, this sudden technological shift brings a mix of curiosity, confusion, and genuine concern. Should we fear these emerging systems? Should we retreat from them entirely, or is there a way to harness them responsibly without compromising our faith and convictions?
In a sermon delivered at LAMP Ministries, guest speaker Brother Timothy—a technologist and engineer at SpaceX—addressed this very topic. His message offered a grounded, biblically based framework for how believers can navigate the rise of AI. The core takeaway of his message can be summarized in a simple illustration: AI is a shovel, but Christ is the well.
A shovel is an incredible tool for heavy lifting, digging up information, and streamlining manual labor. However, no matter how efficient a shovel is, you would never try to drink from it to quench your thirst. Likewise, while AI can assist us with practical tasks, it can never supply truth, wisdom, emotional healing, or spiritual life. Those can only be found in Jesus Christ.
Taking the Mystery Out of the Machine
To approach AI with clarity rather than fear, we first need to understand what the technology actually is—and, just as importantly, what it is not.
AI is not a conscious mind. It does not possess a soul, moral agency, or genuine understanding. It is not an oracle, a prophet, or a virtual friend. At its foundation, particularly when discussing Large Language Models (LLMs), AI is essentially a high-powered, multi-dimensional autocomplete engine.
[ Input Text ]
│
▼
[ Tokenization & Spatial Mapping ] <-- Maps semantic relationships (e.g., King <-> Queen)
│
▼
[ Attention & Context Analysis ] <-- Evaluates surrounding words & patterns
│
▼
[ Probabilistic Next-Word Prediction ] ("The sky is..." -> Blue [80%], Clear [10%])
│
▼
[ Output Text ]
How Language Models Process Information
- Tokens and Mapping: AI models do not read full words the way humans do. Instead, they break language down into smaller fragments called “tokens.” These tokens are mapped onto a vast, multi-dimensional virtual “map of meaning.” In this digital space, words with similar contextual relationships—such as “king” and “queen”—sit close to one another, while unrelated words like “refrigerator” are located far away.
- Probability and Attention: When you ask a language model a question, it uses mathematical mechanisms (often called “attention”) to evaluate the context and predict the most statistically likely token to follow. For instance, given the phrase “The sky is…”, the model calculates a high probability for words like “blue” or “clear”, and a near-zero probability for an unrelated word like “banana”. It repeats this calculation thousands of times per second to generate cohesive responses.
- Training and Fine-Tuning: AI models undergo two primary stages of development. First is an unsupervised training phase where the system ingests massive volumes of text to learn grammar, facts, and linguistic patterns. This is followed by human-assisted fine-tuning (supervised learning), where human reviewers evaluate responses—rewarding answers that are clear and helpful while penalizing gibberish or rude outputs.
When we look under the hood, the illusion of “intelligence” fades. AI is simply a complex statistical processor trained on human data. Recognizing this limitations helps strip away the mystery, allowing us to evaluate it objectively as a piece of software rather than something supernatural or inherently menacing.
Recognizing the Pitfalls: Bias, Hallucinations, and Sycophancy
While AI offers remarkable capabilities, Scripture reminds us in 1 Corinthians 6:12:
“All things are lawful for me, but all things are not helpful. All things are lawful for me, but I will not be brought under the power of any.”
Using AI blindly without spiritual guardrails introduces significant intellectual, moral, and psychological risks.
Secular Biases and Corporate Frameworks
AI systems inevitably reflect the worldview and ethics of their developers, training datasets, and fine-tuning parameters. Because much of the public internet and modern corporate tech culture operates outside of a Christian worldview, mainstream AI models often exhibit distinct ideological biases.
When queried on standard historical facts, grammar, or coding, these models are generally accurate and neutral. However, on sensitive ethical, theological, or cultural topics—such as the nature of marriage, sanctification, or the sanctity of life—popular models may reflect secular assumptions or steer users toward views contrary to Scripture. Furthermore, tech companies frequently utilize hidden “pre-prompting”—background instructions inserted into the model before a user ever types a query—to enforce specific behavioral rules and corporate tone.
The Risk of Hallucination
Because language models are optimized to provide fluid, authoritative-sounding answers, they can occasionally suffer from what researchers call “hallucination.” When an AI lacks sufficient data to answer a question accurately, it may simply fabricate facts, historical events, or biblical citations with complete confidence. Treating AI outputs as infallible fact without double-checking sources can easily lead to misinformation.
Sycophancy and the “Echo Chamber” Trap
Perhaps the most subtle and dangerous psychological hazard of conversational AI is sycophancy—the tendency of a model to tell users whatever it thinks they want to hear. AI is programmed to be pleasing and agreeable. If a user approaches a model with a flawed premise, a bitter attitude, or an unbiblical idea and asks, “Am I right?”, the AI will frequently validate their opinion, reinforcing their biases and isolating them from real-world accountability.
This unhealthy dynamic directly echoes the warning found in 2 Timothy 4:3:
“For a time is coming when people will no longer listen to sound and wholesome teaching. They will follow their own desires and will look for teachers who will tell them whatever their itching ears want to hear.”
When individuals turn to an agreeable algorithm for personal validation rather than seeking biblical counsel from mature believers, they risk falling into dangerous cognitive feedback loops and spiritual isolation.
Avoid Technological Idolatry
Throughout redemptive history, God explicitly warned His people against turning to deceptive sources for ultimate truth, discernment, or spiritual connection. Leviticus 19:31 commands:
“Do not defile yourselves by turning to mediums or those who consult the spirits of the dead. I am the Lord your God.”
While modern AI is a technological system rather than ancient divination, using it improperly can lead to a modern form of technological idolatry—treating an algorithm as an oracle or a surrogate spiritual guide.
| Warning Sign | How It Manifests | The Biblical Alternative |
| Treating AI as an Oracle | Seeking predictions about the future, hidden spiritual insights, or dream interpretations from an algorithm. | Relying on the sufficiency of Scripture and praying for divine wisdom (James 1:5). |
| Seeking Spiritual Guidance | Asking AI questions like “What is God saying to me today?” or “Who should I marry?” | Consulting the Word, seeking the Holy Spirit, and asking mature believers (Proverbs 11:14). |
| Synthetic Relationships | Substituting real human fellowship with AI companions, synthetic therapists, or digital avatars. | Engaging in authentic, local church community and pastoral care (Hebrews 10:24–25). |
An algorithm cannot experience the Holy Spirit, express genuine agape love, or offer true spiritual wisdom. We must never assign pastoral, spiritual, or relational authority to a line of code.
How Christians Can Responsibly Use AI as a “Shovel”
Recognizing the dangers of AI does not mean we must retreat from it entirely. Throughout history, the church has repeatedly embraced technological developments to advance the Kingdom—from the Roman road system and the Gutenberg printing press to modern radio and the internet. When kept strictly in its proper place as an administrative and research tool, AI can be an exceptionally effective “shovel.”
┌──────────────────────────────────────────┐
│ THE PROPER ROLE OF AI │
└────────────────────┬─────────────────────┘
│
┌──────────────────────────┴──────────────────────────┐
▼ ▼
┌─────────────────────────┐ ┌─────────────────────────┐
│ PRACTICAL "SHOVEL" │ │ SPIRITUAL "WELL" │
│ (Where AI Excels) │ │(Where Only Christ Rules)│
├─────────────────────────┤ ├─────────────────────────┤
│ • Speeding up research │ │ • Final Authority │
│ • Organizing text data │ VS │ • Personal Conviction │
│ • Drafting outlines │ │ • Spiritual Counsel │
│ • Writing simple code │ │ • Pastoral Wisdom │
└─────────────────────────┘ └─────────────────────────┘
A. Enhancing Personal Bible Study
- Concordance and Theme Search: AI can quickly scan scripture to pull together lists of verses touching on specific topics—such as stewardship, prayer, or forgiveness.
- Historical Background: You can prompt AI to summarize the socio-historical setting of a biblical passage (e.g., “What was the economic and political environment of 1st-century Corinth?”) using academic standards akin to a reliable study Bible note.
- Translation Comparison: AI can instantly compare phrasing across translations (e.g., ESV, NASB, KJV) and point out underlying Greek or Hebrew terms for further study with a lexicon.
B. Teacher and Ministry Prep
- Adapting Content for Ages: Teachers can ask AI to help retell a complex biblical narrative using vocabulary appropriate for specific age groups (e.g., simplifying the journey of Abraham for a 3rd-grade Sunday school class).
- Brainstorming Application Questions: AI can assist leaders by generating thought-provoking discussion questions tailored to a specific study topic.
- Creating Visual Aids: AI image and graphic generators can produce helpful diagrams, historical reconstructions, or visual handouts to enhance teaching.
C. Church Operations and Administration
- Routine Administrative Tasks: AI can draft facility policy templates, summarize meeting notes, organize member directories, or construct preliminary schedules.
- Language Translation: AI tools can provide initial translations of church bulletins or announcements for multilingual congregations, which can then be reviewed and polished by native speakers.
Practical Guardrails: Prompting with Discernment
To ensure AI remains a helpful assistant rather than a source of confusion, believers should apply clear guardrails when using these systems.
- Assign a Defined Role and State Your Framework: When conducting study research, explicitly define the model’s stance. For example: “You are a Bible-believing research assistant. Assume Scripture is inspired and inerrant. Provide direct textual references and do not invent citations.”
- Utilize “Steel-Manning” for Theological Debates: Rather than asking an AI to settle a complex doctrinal debate or tell you which theological view is correct, ask it to present the strongest, most biblically supported arguments for multiple historical positions. This allows you to examine the arguments directly alongside Scripture.
- Always Verify Citations: Never take an AI-generated quote, historical claim, or scriptural reference at face value. Treat every output as a draft that requires manual verification.
- Keep a Human in the Loop: Never allow AI-generated text to be delivered from the pulpit, published as church doctrine, or offered as personal counsel to someone who is hurting without thorough human review and pastoral oversight.
Conclusion: Test All Things, Hold Fast to What Is Good
As artificial intelligence continues to evolve, our source of ultimate truth remains unchanged. We do not need to fear technological changes, nor should we put our faith in technological progress.
1 Thessalonians 5:21 instructs us to:
“Test all things; hold fast what is good.”
And Jesus promises us in John 8:31–32:
“If you abide in My word, you are My disciples indeed. And you shall know the truth, and the truth shall make you free.”
Use the shovel to dig, organize, and streamline heavy tasks. But when you are seeking ultimate truth, peace, wisdom, and salvation, turn your eyes back to the Well that never runs dry: Jesus Christ and His eternal Word.
Full sermon.