The Real Danger of AI

The real danger of Artificial Intelligence (AI) is not the fact that it will displace workers or that students will cheat on their assignments, although those are serious issues that need to be addressed.

The more profound challenge is the integrity of what we may call “human purpose,” the very reason for our existence. We live in a world dominated by science, technology, engineering and mathematics (STEM), a new era less concerned with the issues that constitute our common humanity.

As is usually the case, Jewish thinkers anticipated something oddly similar. The Golem, an early Frankenstein, is the creation of humans, as Jeremy Dauber points out, which “seeks to imitate the Divine urge” and dates back to the Talmudic period. “Chomer Golem” in Hebrew means “raw material,” something incomplete or unfinished. In each case, the creation of the Golem represents a threat to its creator and is disabled in one way or another. Because it “fails in the most basic acts of humanity,” according to Dauber, it is considered monstrous in the final analysis. The question of whether a Golem can count in a minyan, a prayer quorum of ten Jewish men, is answered by Rabbi Chaim Tzvi with a resounding “No,” as the Golem is seen as a body without a soul.

The best-known Golem was the creation of the famous rabbi, the Maharal of Prague, who, according to legend, infused life into the monster by invoking the ineffable name of God. With its super-strength, the Golem of Prague was enjoined to protect the Jews of his community from pogroms. It, too, became violent in the end and the rabbi was forced to destroy it. As Jay Michaelson writes at the “My Jewish Learning” website, “the power of life is so strong, that it brings both promise and terror.”

It is astonishing to read about human-like machines that challenge humans in the Jewish imagination dating back many centuries. Humans always want to reach out beyond themselves, to achieve what seems beyond human. What seemed possible then, becomes real today with the development of technology. The challenge is to answer the question: What does it mean to be human?

Jews of the past saw the prototype of AI, the Golem, as unfinished or incomplete, potentially dangerous, a tool at best. Modern thinkers recognize the great strides that AI can make in science and health care, among other areas of life, but see its incompleteness in fundamental values such as insight, compassion, justice and wisdom, all unique to humans, whether religious or not.

As Clifford Stoll writes, “Data is not information, information is not knowledge, knowledge is not understanding, understanding is not wisdom.” There is a subtle, gradual development in a person’s life that distinguishes humans from any other species, natural or technological. We have the capacity to learn more than the technical, to examine questions of life lived, to reflect and ponder, and to create art, literature, music that emerge from that process of growth and inner development.

Civil society calls for people to join together in groups for the benefit of one another so that life is enriched for all, whether in churches or synagogues, charities or benevolent undertakings. These are uniquely human impulses.

In “Marginalia” online, Maria Popova writes that “we are each born with a wilderness of possibility within us. Who we become depends on how we tend to our inner garden—what qualities of character and spirit we cultivate to come abloom, what follies we weed out, how much courage we grow to turn away from the root-rot of cynicism and toward the sunshine of life in all its forms: wonder, kindness, openhearted vulnerability.” She expresses the very heart and soul of the human enterprise and thereby demonstrates the limitations of a modern Golem no matter how technically sophisticated.

However, a letter writer to the Globe and Mail adds a disturbing thought that exemplifies the complexity of our modern dilemma: “AI doesn’t need to have a soul, consciousness or emotions. It just needs to be competent. Competence is what we should dread, not malevolence or evil. Anything a human can do, a competent AI could do better and cheaper.” AI won’t be better, but will pass for competent or good because few will know the difference. What percentage of young students do any sustained reading? Ask any high school teacher of English or history.

The problem is not that the machine will take over the world as in a horror movie, but that it will empty people of the richness, depth, complexity, nuance and the uniqueness of life. While we blithely accept what it produces, it will mimic life, but will not be life. Who will know the difference or care?

The problem is not that the machine will take over the world as in a horror movie, but that it will empty people of the richness, depth, complexity, nuance and the uniqueness of life.

The problem, though, is not AI, The Machine, The Golem. It is us. “The fault” wrote Shakespeare, “is not in our stars but in ourselves.” Students don’t read books, they read summaries of books; adults don’t read newspapers and magazines—they get their information from snippets on social media and think they’re informed. It is this new era of self-inflicted ignorance that makes AI dangerous. If you can’t tell the difference between authentic, profound human expression and machine-produced writing, then the fault lies not in the machine but in us.

Only if we reclaim the best of the human experience will we be able to disable the Golem’s power over us.

Dr. Paul Socken is Distinguished Professor Emeritus founder of the Jewish Studies program at the University of Waterloo.