- Rabbi Joshua Franklin told ChatGPT to write him a 1,000-word sermon about intimacy and vulnerability.
- Franklin told his congregation he was "deathly afraid" when applause broke out after his sermon.
A rabbi in New York wrote a 1,000-word sermon using ChatGPT and said the AI might be smart, but it isn't empathetic enough to put preachers out of work.
Rabbi Joshua Franklin was seen in a video dated January 1 telling his congregation at the Jewish Center of the Hamptons that he was reading a "plagiarized" sermon to them. He later revealed it was written by ChatGPT.
"Now, you're clapping — I'm deathly afraid," Franklin told his congregation when they applauded after the sermon. "I thought truck drivers were going to go long before the rabbi, in terms of losing our positions to artificial intelligence."
Franklin told the Associated Press that the ChatGPT sermon was based on a portion of the Torah about intimacy and vulnerability.
Franklin told the AP that his congregation guessed the sermon was written by really "wise, smart, thoughtful individuals." But he added that while ChatGPT may be intelligent, it cannot be compassionate and it cannot love — and so, cannot create the "things that bring us together."
"ChatGPT might be really great at sounding intelligent, but the question is, can it be empathetic? And that, not yet at least, it can't," added Franklin.
Franklin also told the AP he thinks rabbis are "absolutely not obsolete."
"Spirituality often, and the spiritual experiences that we have in life, are often what we'd call ineffable, unexplainable by words," Franklin said.
"And so no matter how good ChatGPT can possibly be at describing, and using language and describing experiences, it can't really understand spirituality. That's a major limitation of it," he added.
Since its release in November, ChatGPT has been used to write a children's book and draft essays, among other things. People are using it to write everything from layoff emails to dating profiles. It's also become a productivity tool at work.
Franklin and representatives at OpenAI did not immediately respond to Insider's requests for comment.