Guy Kawasaki is a marketing specialist, author, and venture capitalist. He’s known for being one of the original Apple employees, responsible for marketing the Macintosh computer line in 1984. If you visit his website, you can ask questions about entrepreneurship, venture capital, marketing, sales, evangelism, technology, writing, and podcasting, and an AI bot will answer based on his writing. It works fairly well.
It got me thinking about how great it would be to do something similar with John Piper’s writings. Imagine training an AI bot on all of John Piper’s books and all the content on Desiring God, and then being able to ask any question without having to wait for him to answer on an Ask Pastor John episode. I think it would probably work well, but I don’t think it will happen. I’m conflicted: I love the idea, and I’m equally concerned by the idea too.
Artificial intelligence is still in its early days. There’s still a lot about AI that we don’t know, and we can only make guesses about what will happen in this field in the future.
Our convictions about AI are a subset of our convictions about technology. Technology is a gift of common grace, but it also has the power to shape us in profound ways. Our responsibility is to use technology responsibly, which involves understanding various technologies as well as we can, and develop the disciplines to use them well.
AI is still a new technology. It’s good at many things, but it’s not completely reliable. Some have compared it to an intern: it can do valuable work, but it requires oversight. Still, it’s good, and it’s getting better and can already outperform humans at many things.
Should we use AI? It depends on what we use it for. Brad Costanzo observes that there are only five things that AI can do: think, create, communicate, analyze, and automate. I don’t have many concerns about using AI to analyze data. I recently used it to compare two insurance quotes and highlight the key differences between them, and the results were helpful. I also don’t have much of a problem with using AI to automate repetitive tasks. I’m in the process of converting old documents into a more readable format, and AI saves me hours. It doesn’t change the content, but it’s really good at reformatting how it’s presented, and I can’t imagine doing the job without it.
My concern is with the three first purposes: using AI to think, create, and communicate. I’m not saying that we should reject it completely in these areas. I spoke to a friend recently who uses AI to help him think through issues, and he treats the technology like a conversation partner. Sometimes AI can help rephrase something that’s we’re trying to say that’s not clear. But I’m also concerned about what we lose when we rely on AI to think, create, and communicate. I want to rely on AI as a tool, not as a substitute for human creativity.
AI can’t worship God. AI can’t treasure God’s word. AI can’t pray. It doesn’t have a soul. AI can’t be grow in God’s grace. AI can’t understand our congregations. There are plenty of things that God made us to do that AI will never be able to do.
AI lacks personhood. Because it’s trained on what people have created, it reflects many of the flaws of fallen humanity. Like all technologies, it will shape us in ways we don’t yet understand, many of which won’t be good for us. We should cultivate hearts that recognize what’s good and that are able to steward new technologies well.
No human technology — including AI — will ever pose a threat to God. But it may pose a threat to us. The fact that we could generate a compelling AI version of Ask Pastor John is not necessarily a reason to do so. Let’s continue to develop a theology of how to use AI in a way that honours God. Developing this theology, by the way, is a task that only humans are able to do.