.png)
Rosh HaShanah 5784
Community Synagogue of Rye
Rabbi Daniel Gropper
In early June I received an interesting text from a colleague. “Were you the inspiration for Debbie Friedman’s Mi Sheberach?” “Huh?”“I’m doing some research on the origins of Debbie’s Mi Sheberach. It says that you were the inspiration for her healing prayer that she wrote in 1985 at Camp Olin-Sang Ruby.” “Who says?” “Chat GPT.”
Just so you all know, Chat GPT was wrong.
As flattered as I was by this ChatBot, I’ve never been to OSRUI. In 1985 I was trying not to flunk physics and was trying to get Jacolyn Cutler to notice me so I could ask her to Jr. Prom. I managed to pass physics - barely. Jacolyn went with someone else.
I knew that the topic of AI was eventually going to make it into a sermon, but as it’s coming so fast, faster than any of us can imagine, I’ve chosen this day of new beginnings to start the conversation, and to consider how Judaism in particular and religion in general might respond.
The first thing we can do is learn from it. There is a Hasidic tale where Reb Avraham Yaakov says that it is possible to learn great truths from inanimate things. That once we see the world as abounding with messages from God, everything can be a spiritual teacher. From a train, we learn that in a single second, we can miss the whole thing. From a telephone, we learn that what we say here can surely be heard there. What then can we learn from A.I.? That what a single person can know is like a grain of sand but combined, our knowledge can be like the sands on the shore.
The second thing we can say is that the Hebrew is better than the English. The Hebrew for Artificial Intelligence is - בִּינָה מְלָאכוּתִית. A more literal translation would be artificial understanding or artificial comprehension as Bina means understanding or comprehension. The word usually used for intelligence is חכמה which is more closely related to wisdom. But since wisdom is “metabolized experience which leads to distilled compassion,[1]” and AI can have neither metabolized experience nor compassion that is distilled - the Hebrew term is more fitting - artificial understanding or artificial comprehension; yet for our sake, let’s just call it AI.
And the third thing we can say - for the most part - is that while it will alter our existence, it won’t render us lamblike as Dave was to the HAL 9000 in Kubrick's Space Odyssey; or victims to machines that require us as energy sources as in The Matrix; or lazy layabouts like in Wall-E. Instead, just as Isadore Rabi’s mother asked the Nobel laureate physicist upon his return from school if he asked a good question, AI will require us to think harder and to ask better questions, to be more present and mindful and to be more decisive so that we make the ultimate decisions and not leave it to an intelligence that is artificial.
As I’m sure you’ve read, talked about or heard, the launch of Chat GPT, the first really convincing AI chatbot last November sparked widespread excitement and fear.
Teachers, journalists, screenwriters and disinformation experts started asking questions like: What if it helps students cheat? Could it replace me? Will it amplify conspiracy theories? Chat GPT could answer complicated questions, draft emails, plan custom vacations and much more.
In February, Microsoft unveiled Bing, the chatbot that both delighted and disturbed users with eerie interactions. Now we’ve got GPT-4 — a multimodal large language model that can respond to text as well as images. We’ve learned of people who use ChatGPT to provide language to have more productive conversations with their teenagers, draft love letters, negotiate their rent, understand medical results and affirm their gender identity.
The reality is that different versions of AI have been around so long that it’s become an invisible part of our lives. Do you use Alexa or Siri? Has an online assistant popped up to help you? Does your word processing software suggest words to finish your sentence? You’ve interacted with A.I. But this time it’s different. So what changed?
About 10 years ago, the tech industry began writing code that shifted from telling a computer what to do, to telling a computer what to learn what to do. These A.I. systems started learning to put language together in the way you and I do, by predicting what words come next - like the auto completion on your email but taken to an enormous scale. Chat GPT explained it to me this way, “Think of ChatGPT's development like learning a language. Initially, it's like a baby babbling, making sounds without understanding. As it gets more data (like a child hearing more speech), it learns patterns and starts forming coherent sentences. Over time, it becomes better at understanding and responding to conversations, much like how humans develop language skills through exposure and practice.”
The difference here is that machines can learn in ways that humans cannot and they are becoming faster and more powerful. AI pioneer Geoffrey Hinton explains it this way, “I have a brain that can learn a certain amount of information. After I learn it, I can convey it to you. But that’s a slow process. Imagine that there are a million people and, as soon as one person learns something, they all know it. [The way these networks work is that], a small network that can learn a little information can be connected to all the other networks that have learned from other parts of the network. They can all learn in tandem and can trade what they've learned with each other in an instant. It means that a digital agent can read the entire internet in a month. We can’t do that. We can’t trade what we’ve learned so easily with each other.”[2]
I know it sounds like a book or movie plot - but it’s not. What is happening with these neural networks is the biggest technological advancement since the industrial revolution, electricity, or the wheel. Some have even compared it to Promethean fire. As Danny Schiff writes in Judaism in a Digital Age, “In the coming years, Artificial Intelligence will reshape the way we think, it will fundamentally, perhaps irrevocably change the very nature of human existence and it will do so at a speed faster than we ourselves will be able to think about.” Let’s ponder then, for a few moments, how we might respond.
In 1949 at the start of the computer age, American mathematician and philosopher Norbert Wiener said “If we move in the direction of making machines that learn and whose behavior is modified by experience, we must face the fact that every degree of independence we give the machine is a degree of possible defiance of our wishes.” Unless we unplug all the computers, this technology is here to stay and it will only grow more intelligent.
As hard as change might be for some - myself included - we have an opportunity to harness it for good and to embrace it. It’s how we responded to the matzah making machine in 1859. Oh, you didn’t know there was a matzah making machine? It was one of the great technological advances of the 19th century…the electric lightbulb, the telephone, and machine made Matzah. Even though those who objected to it said that machine made matzah was matzah without a soul and that “matzah made with a machine was no better than a loaf of bread,”
We embraced the change. Even though the Bible shows little interest in timekeeping, we responded to sundials, clocks, and other timekeeping devices and embraced them to become a tradition that seems obsessed with time. If you don’t believe me, just look at a Jewish calendar. Shabbat candle lighting times are so important they are printed IN RED!
I am sure that when the first person discovered that they could hollow out a ram’s horn to make a sound, it frustrated the fire makers who used to signal important information by lighting fires on mountain peaks. A shofar was portable. It required no fuel. It could be used in any weather. As long as you were killing the ram for food or sacrificing it as a tribute to God,
the ram’s horn was a bonus. It was free. Of course, the fire makers would protest because they saw their jobs being threatened but once that first shofar was sounded, there was no going back. In this fictitious debate though, I imagine the Jewish response. Why should it be either/or? Can’t we find a third way? A both/and? I imagine that an agreement was reached between the shofar blowers and the fire makersthat to announce news, such as a new month or a new year, the shofar blower would walk to a designated spot, blow the shofar and then, upon hearing it, the fire makers would light signal fires throughout the land. Maybe other shofar blowers would join them. Just as Judaism found a way for shofar blowers and fire makers to coexist, thinkers in the area of AI are saying that the future is not AI vs. humanity but AI and humanity, working together to make our world a better place.
One organization, for example, is using AI to identify faces in Holocaust photo archives and to compare it across collections so that users can discover loved ones and relatives lost in the Shoah. Another works with Holocaust survivors to generate images that visually represent their memories of surviving the genocide. And researchers at NYU’s Grossman School of Medicine and the University of Toronto designed an AI program that could allow scientists to treat Jewish genetic diseases through gene-editing methods.
As AI reduces the need for human labor, the question of how to spend our newfound leisure time becomes central. Alon HaLevi, director of Facebook AI notes, “I like to think about AI as a set of technologies that will enhance our human capabilities rather than replace us. I would like an AI that makes everything I do during the day easier and enables me to focus on the essence of what I’m doing, on the fun part of the work.” Here, the Jewish tradition of adult study of Torah could serve as a model for engaging in intellectually and spiritually fulfilling activities. This concept of a community beit midrash could serve as a valuable template for societies grappling with the challenge of how to cultivate a sense of purpose in an era of increased free time. Jewish tradition also provides a framework for considering the place of technology in our lives. The concept of Shabbat, offers a model for maintaining a healthy balance between technological advancement and human well-being.
One of the greatest things Judaism gives us is a notion of boundaries. Judaism says, you can’t eat everything you like, you can’t have sex whenever you want, you can’t treat your workers any way you like, you can’t amass as much wealth as you want, you can’t take other people’s ideas and claim them as your own, you have to care for this planet, even if it costs more; you have to respect your elderly, even if they haven’t specifically done anything for you, and you can’t be greedy. We might not always agree with these rules, we might not always follow them but the boundaries exist to keep us on a path or to help us find the path when we are lost. This notion of boundaries can be applied when thinking of some of the issues raised by AI.
One example is the issue of privacy. As AI increases demand for the data gleaned from facial recognition technologies, we have to say that there’s something fundamentally exploitative about not respecting humans as unique and worthy individuals made in the Divine image but instead, as objects to be wrung. Given how we were reduced to numbers in the Shoah, there’s a special Jewish sensitivity to reducing people to 1’s and 0’s and using them as data fodder in a surveillance capitalist ecosystem. What should the limits and boundaries be when it comes to protecting our privacy? Judaism has what to say on this and on many moral topics.
At its core, what AI does is to make us rethink what it means to be human. And when we have to rethink what it means to be human, there is then a place to discuss them in religious terms. David Zvi Kalman of the Shalom Hartman Institute points out that human beings are, within the Jewish tradition, literally an “artificial intelligence” created by God. The first golem—a figure later strongly associated with machinery and computing—is the body of Adam, that first human, right before God breathes life into it. In other words, the story of AI is not that different from our own, except now we’re the ones doing the creating. What happens when the created becomes the creator?
To that very question, Rabbi Danny Nevins argues that the advent of machines that generate ideas and text and even speech should not diminish the value of humanity, but rather cause us to look deeper at the true foundation of our worth. As he says, “Humans should be valued not only for their ability to generate novel content, but for their relationship to each other, to morality, and ultimately to the divine source of their lives.”
The relationships we have with each other, the relationships we have with God, the fact that we are not just human doings but human beings, that we can love, that we can dream, that we can have creative new insights for previously unanswered questions or new spiritual challenges; that we can miss the mark and try to make it better, that we can forgive and sometimes forget, that we have souls, that we can die, that we can physically reach out and touch someone and comfort with a hug, not just words; this and more makes us human and distinguishes us from AI.
Philosopher and media theorist Marshall McLuhan framed it by saying that technology is an extension of ourselves. Just as God led Adam and Eve around the Garden saying, “this world is yours, make it better.” AI has the potential for us to improve who we are as human beings, chief of which is learning how to become more creative, to use our imaginations more fully.
Chat GPT and other generative AI’s may allow us to look up anything without having to really think about what the answer is. It may generate the draft of a speech or the outline to an essay but that isn’t the same as the creative process of using our imaginations to generate thoughts. The tablets that ultimately wound up in the ark were not the ones God originally gave Moses on Mount Sinai. The ones that ended up in the ark were the ones Moses carved himself. The Torah teaches us over and over that God wants us to use our minds, to be creative, to put in the effort, not to take the easy path. At the very least, generative AI should push us, as Isadore Rabi’s mom did with him, to learn how to ask better and better questions.
Rabbi Yitzchok Adlerstein, an Orthodox rabbi from Los Angeles believes that through AI, people will rediscover the soul. “I’m reminded of the Hasidic version of Descartes,” he says. “A person came to his rebbe and said, ‘Rebbe, how do I know that what I’m going through is real? Maybe I’m just dreaming and there is no reality.’ The rebbe did not say ‘Cogito ergo sum.’ He said, ‘Tell me, do you ever have thoughts of teshuvah?’ The man said yes. The rebbe said, ‘In that case, you’re not living in an imaginary world, because in an imaginary world, people don’t think of doing teshuvah.’ The impetus for teshuvah comes from the depths of a person’s soul, which is deep inside us and comes from God. AI can’t really do teshuvah. It can’t feel regret and it can’t try to reconnect with its ultimate source, that mystery we call God that inspires us to do better, to be better. Only humans can do this - I know this because I asked Chat GPT this question - and I think AI will force people to take a closer look at that.
There are so many questions AI is raising, questions I don’t want to leave here in this sanctuary after the final shofar blast carries our prayers into the universe. That is why this fall and winter we will hold a series of conversations on this topic. Led by experts on AI who thankfully are members of our congregation, we will gather to discuss the economic, social, and religious implications of this new technology.
As a temple community we will look at modern dilemmas raised by AI through a lens of ancient texts and our own lived experience, to make choices based on knowledge, just as our Reform Jewish tradition has always done. Maybe our learning will even inform the future of AI? (a rabbi can dream).
Author and educator Parker Palmer tells the story of a weeklong class with Outward Bound, during which he was given a rope and instructions to rappel down the face of a cliff. He got to flailing, fast—frozen in place by the sight of the drop beneath and unable to listen to the commands being called to him from the safe ground above. When things came to a stalemate, his instructor reminded him of the motto of Outward Bound: "If you can't get out of it, get into it."
When it comes to AI, this is the way the world is going, we can’t get out of it. We have to get into it. The way we get into it is by using the gift God gave us - our knowledge, wisdom, and our understanding of what’s real and what’s not, of fact-checking and not taking things at face value. This is the key skill we need to become better at as humans. And I think we’re going to be fine. If there is one skill we as Jews have honed over time, it’s the ability to question everything. It’s our super power. So let us dive in, let us discover, and let us discern what it really means to be human, the ability to ask a good question - something God has modeled for us since the dawn of time when God asked Adam, “Ayeika, where are you?”
[1] As Defined by Chip Conley, https://mastersofscale.com/the-rise-of-the-wisdom-worker-chip-conley/ [2] NY Times The Daily Podcast, May 30, 2023