How do we prevent the worst-case AI scenario? Religious AI scientists
AI technologists need to view a higher power as the ultimate creator, which would temper the scope of their inventions
Photo by Bahadir Eroglu/iStock
Micah Ezekiel May 23, 2023
Lots of people are scared of artificial intelligence. Television writers are on strike in part out of fear that AI will replace them. Sam Altman, the CEO of OpenAI and a leading AI software developer, asked Congress to regulate the technology in a recent testimony, warning that AI “can go quite wrong.”
These concerns are part of a fierce debate raging about how grim the future of artificial intelligence will be. Eliezer Yudkowsky, a founding father of AI, worries that without severe regulation, a human-like AI, called artificial general intelligence, will eventually kill us all.
Tyler Cowen, an economist and blogger, demurs, saying that Yudkowsky’s certainty in the worst-case-scenario AGI (artificial general intelligence) outcome is overblown. Because we don’t know how humans will ultimately use AI, Cowen argues “radical agnosticism is the correct response, where all specific scenarios are pretty unlikely.” The only thing for certain, says Cowen, is that AI will alter history and cause turbulence in people’s lives. We just don’t know the specifics.
If AI scientists were properly spiritual, properly religious — not necessarily observant, but had a religious mindset — then humanity as a whole wouldn’t have much to worry about. Infusing religion into technologists, especially AI scientists, might help tame temptations to create the worst-case-scenario outcomes of new technology. If they had faith in God, we might be OK.
As a student of natural-resource management, I’m primarily concerned with how technological progress affects nature, but as a believer in God educated in Jewish texts and ideas, I’m also concerned with how new technologies like AI will shape our societies psychologically and spiritually.
Creating history-altering technologies, or anything at all, is a God-like behavior. Humanity acts as gods when it creates. In the monotheist’s mind, God created the Heavens and the Earth from nothing. Humans create things from seemingly nothing too.
Today, we are more God-like than ever, according to St. Augustine’s description of God as omnipotent, omnipresent and omniscient. (I admit, I used ChatGPT to figure out who coined the “three omni-” description of God when Google proved difficult.) Humans have dominion over the earth and the heavens through air and space travel; omnipotence. We are nearly omnipresent through communications technology. And with the internet, search engines and now AI, human beings are nearly omniscient. We’re only going to know more as history marches on.
It’s not a bad thing that we are creators. The Torah knows this and tells us this in its first chapter: “And God created humankind in the divine image, creating it in the image of God.” Since God commanded humans to “fill the earth and master it,” humans have been creators. In Genesis, we see the first humans work the land as agriculturalists, then later build towers and cities, iterating upon and changing God’s initial creation. In short, God created creators.
I’m obviously not the first Jew to have this insight. It’s the entire point of Rabbi Joseph B. Soloveitchik’s book The Lonely Man of Faith, in which he explains two archetypes of what it means to be human based on the first two chapters of Genesis. His Adam I archetype, derived from his reading of Genesis I, is the Creator, one that finds dignity by becoming as close to God-like as possible, discovering scientific principles and using them to create technological advancements in areas like medicine and travel.
But Soloveitchik also sees a complement to human creativity present in Genesis, a second archetype he calls Adam II. In the second chapter of Genesis, God places Adam in the Garden of Eden “to till it and tend it.” He is the human that needs to make meaning out of Creation, not things. Instead of creating scientific laws to find out how the world works, Adam II looks at the world qualitatively and asks why it is the way it is. He has a religious mindset that counterbalances the creative impulses of Adam I. Both archetypes are integral to the Jewish conception of what humans are put on earth to do: to not only create but make meaning out of our creations.
The siddur also offers us daily reminders of a religious counterbalance to our creative impulses. The late Rabbi Jonathan Sacks teaches that the siddur’s narrative, repeated in our tefillah every day, goes from creation, to revelation, to redemption, and from love to awe. If those seeking to create an artificial general intelligence go beyond the first step of creation, as Rabbi Sacks suggests, they would see there is more to being human than creating for the sake of creation. With a religious mindset, AGI developers might think long and hard before giving their “robots” unfettered abilities that could wreak havoc on flesh-and-blood humans. Instead, they might see the goal of AGI as an aid to a higher redemption.
AI technologists need to view a higher power as the ultimate creator. The Shulchan Aruch — a codification of Jewish law followed to this day by many Jews — recommends thinking of God upon waking up: “One should strengthen himself like a lion to get up in the morning to serve his Creator, so that it is he [the human] who awakens the dawn.”
Our society has no trouble awakening the dawn, but we don’t realize that we serve a Creator in order that we might create. By knowing that there is a higher creative power than them, AI developers would humbly resist the urge to create systems that could wipe out humanity because they would know they have no right as humans to do so.
I don’t know what religious beliefs (if any) leading AI researchers hold today. But I feel confident that if they shared some of the divine awe and appreciation for our world, and our collective luck at being alive here on this planet, then their AI research would proceed from a place of nuance, humility and respect for Earth.
Cowen is probably right — we don’t know what a human-like AI will do for, or to, us. But instead of asking the questions what can we create or how will we create it, AI scientists can ask themselves a fundamentally religious, ethical and spiritual question: Why are we creating? In posing this essential question, AI developers might not only understand that their creations could serve a higher purpose. It may also protect the world from AI technology going horribly wrong.
To contact the author, email [email protected].
A message from our CEO & publisher Rachel Fishman Feddersen
I hope you appreciated this article. Before you go, I’d like to ask you to please support the Forward’s award-winning, nonprofit journalism during this critical time.
We’ve set a goal to raise $260,000 by December 31. That’s an ambitious goal, but one that will give us the resources we need to invest in the high quality news, opinion, analysis and cultural coverage that isn’t available anywhere else.
If you feel inspired to make an impact, now is the time to give something back. Join us as a member at your most generous level.
— Rachel Fishman Feddersen, Publisher and CEO
