The New York Times called it "the best artificial intelligence chatbot ever released to the general public." The Atlantic called it “the generative-AI eruption" that "may change our mind about how we work, how we think, and what human creativity really is." A high school teacher called it "the end of high school English." A New York rabbi saw it as the higher power that gave him his latest sermon. The Who might say this about it:
Then I'll get on my knees and pray
We don't get fooled again
Meet the new boss
Same as the old boss
(from “Won’t Get Fooled Again” by The Who)
If you haven’t figured “it” out yet, “it” is ChatGPT (Chat Generative Pre-trained Transformer), the chatbot from OpenAI which in just three months has some people hailing it as the most human text-to-speech AI app ever, while others fear it is the end of education, jobs, and perhaps even religion. Released as a free program by OpenAI in November 2022 as part of its GPT-3 family of large language models, ChatGPT was immediately picked up by the general public which marveled at its ability to quickly proved detailed responses and coherent answers across many domains of knowledge with what seemed to be pinpoint accuracy. As is the case with so much new technology, Chat GPT was speedily applied for nefarious purposes: generating fake resumes, phishing and scamming, writing fake articles and term papers, fooling dating apps, writing erotica and even creating malware. To the relief of many, its factual accuracy is often not that great … yet. However, it is clear ChatGPT could be on its way to becoming the new writer, new help desk worker, new boss … and perhaps even a new god. Can the congregation be fooled?
“Despite the hardships that he has endured, Joseph is able to find in his heart the ability to forgive his brothers for their past wrongdoings. By approaching them with openness and vulnerability, he is able to heal old wounds and to create deeper, more meaningful bonds with his siblings.”
That is part of a sermon delivered recently by Rabbi Joshua Franklin, of the Jewish Center of the Hamptons. For those not familiar with it, it refers to the Torah portion Vayiggash. Vayiggash and the Old Testament Book of Genesis tell of the journey of Jacob to Egypt with his sons, one of which is Joseph – the owner of the famous ‘coat of many colors’. In the homily, Rabbi Franklin talked about the importance of intimacy in building relationships, described vulnerability as “the willingness to show up, and be seen when we have no control over the outcome,” and quoted Professor Brene Brown, who is nationally known for her research on shame, vulnerability, and leadership. The rabbi closed the sermon with:
“May we be blessed with the courage and strength to be open and vulnerable.”
He then turned to the congregation and, in an open and vulnerable way, admitted that the sermon had been plagiarized. He then asked them to remember the words and attempt to figure out who the author was. Their responses included other local rabbis, Franklin's father, and even the late Chief Rabbi of the United Hebrew Congregations of the Commonwealth, Jonathan Sacks. After no one guessed the source, Rabbi Franklin admitted – again, in an open and vulnerable way - that the homily had been written by ChatGPT. As he looked out at their shocked faces, the rabbi decided he’d better explain what he did … and why.
“I gave it a prompt, I said write the sermon, in a voice of a rabbi, of about 1,000 words, connecting the Torah portion this week, Vayigash, to the idea of intimacy and vulnerability, and quote Brene Brown.”
Thinking back, the congregation realized that this is exactly what Rabbi Franklin/ChatGPT talked about in the homily. The Times of Israel reports that Franklin then got a response from the members of the temple that he wasn’t expecting.
“Now, you’re clapping, I’m deathly afraid. I thought truck drivers were gonna go long before the rabbi, in terms of losing our positions to artificial intelligence and self-driving vehicles.”
While that comment probably insulted the truckers (would ChatGPT have said that?), Franklin was more concerned about the safety of his own job and the state of his – and all other – religions. On “The Claman Countdown” talk show, he revealed that the first thing to scare him was that ChatGPT was “able to synthesize two separate ideas” – the story of Joseph and the idea of vulnerability. It quickly responded with a 1,000-word sermon “that's not just cogent, but actually pretty strong and articulate.” The second thing that scared him was that they were impressed and didn’t catch any errors.
“Look if I read that, I would know the various faux pas. There are distinctive markers that aren't how many rabbis would speak.”
Finally, it was true confession time - he admitted that he really wasn’t worried about losing his job because he felt that ChatGPT was incapable of empathy – it didn’t have a ‘soul’, love or compassion. On the other hand, he admitted that the world we live in today “is driven by data and driven by information.” The sermon passed the smell test of his data-driven congregation. Could it be that ChatGPT is what the people want? He pointed to the one area where it would fall flat on its face – a non-empathetic ChatGPT rabbi creating sermons exclusively using Internet content could easily include something antisemitic.
“I don't think ChatGPT or any kind of artificial intelligence will replace us, but it will push us. It'll force us to evolve in what we do and what we do best.”
Rabbi Franklin promised his flock that he wouldn’t plagiarize a ChatGPT sermon again, but admits he sees a benefit to it – competition. In that respect, he may be right. A few years ago, robot priests were the big new idea in the marriage of tech and religion. A robot priest called BlessU-2 developed in Germany delivered blessings in five languages and beamed light from its hands – not exactly the same as an ‘original’ sermon. A robot priest named Mindar was unveiled at Kodaiji, a 400-year-old Buddhist temple in Kyoto, Japan, where it delivered pre-recorded prerecorded sermons. At the time, its developers promised that machine-learning capabilities would one day help Mindar to customize feedback to the specific spiritual and ethical problems of humans.
If a savvy religious techie merged BLessU-2 and Mindar with ChatGPT, would it end the need for modern religion? That techie better watch out for lightning bolts.
Or was The Who singing about this very day? Will humans not get fooled again by a ChatGPT sermon when they get down on their knees and pray?
Or is this new boss same as the old boss?