AI Doesn’t Mimic God’s Intelligence
The Gospel Coalition
TGC supports the church by providing resources that are trusted and timely, winsome and wise, and centered on the gospel
Artificial intelligence has become impressive. Large language models (LLMs) such as ChatGPT can now mimic human experts. It’s intimidating to think about a new world where computers sound as smart as a practiced hospital nurse or legal assistant.
ChatGPT works by absorbing a vast amount of literature—more than any person could ever read. As these models improve their ability to parrot human writing, the LLMs will keep sounding smarter. AI agents could potentially learn something humans don’t yet know by running new scientific experiments. But even that would mean they’re only uncovering another corner of the universe God created and already knows intimately.
While some have compared LLM breakthroughs to “building god,” or used descriptors like “god-like” to describe what AI can do, the distance between even the most advanced LLMs and God remains infinitely vast. AI can mimic human intelligence, subject to constraints (like electricity usage). However, even if AI tools became leading poets or groundbreaking scientists—mimicking the brilliance of human creativity—this wouldn’t put AI in the same class as God.
Difference Between Human and Divine Intelligence
The Bible’s ancient writers went out of their way to differentiate human intelligence from God’s knowledge.
Scripture repeatedly emphasizes God’s omniscience, portraying him as the possessor of infinite knowledge and wisdom. Psalm 147:5 declares, “Great is our Lord, and abundant in power; his understanding is beyond measure.” This limitless understanding is beyond human comprehension, as Isaiah 55:9 illustrates: “As the heavens are higher than the earth, so are my ways higher than your ways and my thoughts than your thoughts.” God’s intelligence transcends time and space, as Romans 11:33 exclaims: “Oh, the depth of the riches of the wisdom and knowledge of God! How unsearchable are his judgments and how inscrutable his ways!”
While humans possess finite intelligence—prone to error and limited by the confines of the physical world—God’s intelligence is characterized by perfection, infallibility, and an eternal perspective.
Uses and Limits of AI
Mark Zuckerberg recently said about AI, “There’s all this science fiction about creating intelligence where it starts to take on all these human-like behaviors. . . . The current incarnation of all this stuff feels like it’s going in a direction where intelligence can be pretty separated from consciousness, agency, and things like that, which I think just makes it a super valuable tool.”
The type of work performed by AI can be useful to a human organization (even a church). I’m optimistic LLMs can be tools that help humanity solve challenges. The Bible says discovering new ideas is “the glory of kings” (Prov. 25:2). Nothing indicates that having more intelligence is bad if it guides people to truth.
With several researchers from Samford University, I’ve demonstrated that AI can perform well; however, it sometimes produces falsehoods. Our research shows that if you ask ChatGPT to create academic reference citations, the program may invent names of books that don’t exist. Additionally, our findings suggest the reliability of the ChatGPT model decreases as the prompts become more specific.
Christians shouldn’t fear AI. But we should be prudent in how we use it. Having a machine that can answer any question puts us in an interesting position. Do we know how to craft the right questions to pose to AI (“prompt engineering”)? Will we build talking machines that flatter us and help us commit crimes, or will we progress in understanding through AI’s ability to quickly synthesize vast amounts of information? Is “intelligence” that’s more informed than us valuable if it isn’t accompanied by discernment and moral evaluation? Without spiritual discernment and wisdom to guide this process, humans could end up more confused than ever.
King Solomon tried wisdom, along with everything else a human could pursue in this world. He reported, “With much wisdom comes much sorrow; the more knowledge, the more grief” (Eccl. 1:18, NIV). Having access to vast amounts of rapidly synthesized information doesn’t insulate us against suffering. Christians should neither fear LLMs nor count on them to solve all our problems.
Comfort of God’s Superior Wisdom
God’s higher intelligence comforts us in a changing world. His omniscience assures believers he comprehends the complexities of every situation, even amid our uncertainty. It’s reassuring to know God sees beyond the present moment and understands the broader implications of events. He also cares about individuals. Psalm 139:1–2 affirms this: “O LORD, you have searched me and known me! You know when I sit down and when I rise up; you discern my thoughts from afar.”
God’s wisdom surpasses human understanding, providing guidance in times of upheaval. Proverbs 3:5–6 encourages trust in the Lord’s wisdom: “Trust in the LORD with all your heart, and do not lean on your own understanding. In all your ways acknowledge him, and he will make straight your paths.” Believers find solace in relying on God’s superior wisdom, knowing his plans are ultimately for their good (Jer. 29:11; Rom. 8:28).
It may be hard to fathom how far technological and scientific understanding will eventually progress in the age of AI and in whatever age comes after. God’s unchanging nature offers stability and security in a changing world. Hebrews 13:8 declares, “Jesus Christ is the same yesterday and today and forever.” His character and steadfast love are consistent, regardless of whether the world is rocked by the printing press or social media. Technology has advanced, but God is the same.
Joy Buchanan is an economics professor at Samford University in Birmingham, Alabama. Follow her on Twitter or read her blog.
Software Engineer @ Compassion
4 个月Misunderstanding, hyperbole, and anthropomorphism seem common when it comes to perception of LLMs. In my opinion, industry standardization of human words like “intelligence”, “reasoning”, “inference”, etc doesn’t help. I try to communicate to others whenever I can that — at a fundamental level, as is the case with all computing — LLMs do math. LLMs accept input and return output based on “training” ( read programmatic assignment of statistical probabibility ). They neither reason, nor lie, nor prophesy. LLMs are a new riff on software/hardware that can help. That’s it! :)