The AI Generated Sermon

Photo by ThisIsEngineering on Pexels.com

Recently someone made a confession:

  • They now used ChatGPT to generate all their sermons
  • The congregation were happy and gave favourable feedback on these AI generated talks
  • It freed up their time to focus on pastoral visits.

I strongly suspect it is a spoof/satirical comment, intended as satire but it could easily be genuine.  After all, I’ve recently been writing about whether ro not we should borrow sermons andis  illustrations from others. The person concerned would only be going one step further in making use of AI.

ChatGPT is an Artificial Intelligence tool, essentially working like an advanced database. It has collated a lot of information and knowledge available on the web, enabling you to get quick answers to questions.  One of our church members recently tried it out to see how it would answer our usual midweek Life Group question “What is going on in the passage?”  The Bible passage in question was Isaiah 35 and here is what ChatGPT came up:

Isaiah 35 is a chapter in the book of Isaiah in the Old Testament of the Bible. The chapter describes a prophecy about the future restoration of the land of Israel and the joy that will come to its inhabitants.

The chapter starts with a description of the wilderness and the desert becoming a place of beauty and abundance. It says that the desert will bloom like a rose and be full of verdant vegetation, such as cedars, myrtles, and pines. The prophecy also describes that the land will be watered with streams and springs and that the highway of the Lord will be made straight.

Isaiah 35 goes on to describe the deliverance that will come to the people of Israel. The eyes of the blind will be opened, the ears of the deaf will be unstopped, and the lame will leap like a deer. The chapter also speaks of the restoration of the exiles to Zion and the celebration that will ensue.

Overall, Isaiah 35 is a message of hope for the future of Israel and the joy that will come to its people. The prophecy foretold the restoration of the land, the healing of the afflicted, and the ultimate victory of God’s people over their enemies.

It’s not a bad answer is it? In fact, we might consider it a fairly solid exposition of the Bible passage.  However, I want to suggest that there are two crucial things missing from it.  First, it fails to look back in Israelite history and see the all important echoes of the Exodus when God delivered the people out of Egypt and the hand of Pharoah.  God was promising the people of Judah a similar rescue event when he would bring them safely back from Babylon.  Secondly, it fails to take us as believers to where we need to go if we are to apply the Bible passage properly. It doesn’t take us to Christ and through him to our life circumstances as believers.

Now, perhaps there, we have just uncovered the limits and the downside of AI.  However, if we are honest, those of us who are preachers have found ourselves preaching sermons that would not sound that much dissimilar to the exposition here. And if that’s the kind of thing we regularly preach, then our congregations won’t be able to tell the difference when we pluck together borrowed bits of other sermons and quotes from books or even if we simply regurgitate the results from an AI inquiry.

Indeed, if we do, then our churches might do well to cut out the middle man and simply get the AI device installed with voice recognition in the auditorium.  They could probably get it to do home pastoral visits too.

This is why it’s important that those of us with preaching and pastoring responsibilities take them seriously.  We will really want to get to grips with what is going on in the text, to see what it is saying, to understand both the echoes back to God’s saving acts of the past and the foreshadowing forward to Christ.  It also means that we will want to get to know our congregations well and know what is happening in their lives.  We will want to apply God’s Word accurately and lovingly to their lives. AI cannot do that.

%d bloggers like this: