AI has revived ancient questions: What becomes of human beings when something greater than us can do, know, and judge better than we can? If machines produce abundance without labor, who should share in it? If they make knowledge instantly accessible, what becomes of learning? If they reason better than we can, what calls must remain human?
The technology is new. The predicaments are not.
Thousands of years before AI, Homer envisioned golden handmaidens with “intelligence, speech, and strength”; Aristotle imagined self-operating tools; Greek myth gave us Talos, an autonomous weapon; Talmudic scholars fabricated synthetic humans; and ancient civilizations consulted super-intelligences. Our world is not theirs. But our deepest questions are.
Start with abundance. AI promises plenty on a scale the ancients could only mythologize, reviving two old dilemmas: Who is that abundance for, and what is it for?
On the first question, their answer was expansive: Abundance was for everyone. In the 12th century, Maimonides predicted a future in which “goods will flow in abundance, and all the delights will be freely available as dust,” where “little effort will yield great results,” and there would be “no famine, war, jealousy, or competition.”
Three centuries later, Thomas More’s Utopia (whence the word) envisioned “a great abundance of all things,” so broadly shared that “no man can want or be obliged to beg.”
Yet left to market forces, AI’s plenty will likely accrue to the few who own the technology, not the many whose livelihoods it disrupts. If abundance is to flow universally, we will need to design the economic plumbing to make it so.
On the second question, what abundance is for, the ancients had a clear answer: Learning. The rabbis idealized Torah lishmah, study not for career or status but as an end in itself. The Greek word schole, root of “school,” meant leisure, framing study as an alternative to work, not a prerequisite for it. Both traditions saw the deepest human vocation not as production but understanding.
While AI can be the consummate teacher, it also undermines some incentives to learn. We have long sold education as a route to livelihood; that calculus is looking shaky when no amount of study will make us smarter than an LLM (large language model). Nor does easy access to information guarantee understanding.
In the Phaedrus, Plato warned that writing would leave people able to “appear to know much while for the most part they know nothing,” with “the show of wisdom without the reality.” AI sharpens the danger. If we let it do our writing, reasoning, researching, and deciding, the faculties that make thinking possible may atrophy. Having access to the truth is not the same as knowing the truth.
So we must recover learning as an end in itself. Maimonides put the endpoint simply – study “not for money, not for honor, but because knowledge is worth pursuing for its own sake.”
Beyond learning
Yet much of humanity will need more than schole. Rabbi Yosei Hagalili taught that “a person dies from idleness,” and Rabbi Yehuda ben Bathyra taught that one without work should find “a ruined courtyard or a ruined field, and go and occupy himself with it.” In an age of plenty, the complement to improving our minds will be improving our world.
The Hebrew word mitzvah means both “commandment” and “good deed.” A life built around knowledge, structured time, and the turn toward others is not a consolation prize for a world without work. It is what the abundant era was supposed to make possible.
For the ancients, a world of plenty was largely a thought experiment. Oracles, on the other hand, were lived experiences.
Biblical Israel consulted the Urim and Thummim on questions of national consequence. The Greeks built a civilization around Delphi, where the Pythia settled disputes, authorized wars, and shaped policy. To their adherents, these were authoritative voices from beyond ordinary human judgment, so both faced the same problem: If the superior intelligence has spoken, what room remains for human judgment?
Delphi embodied one answer: Divine authority was real, if open to interpretation. Its most famous inscription was not a prophecy but a command: “Know thyself.” Its deepest function was not to deliver answers but to force self-examination.
AI presents the inverse problem. Oracles were Delphic, and their ambiguity invited human interpretation. AI is increasingly clear, confident, and accurate enough to invite submission. If AI diagnoses cancer better than any radiologist, adjudicates disputes more fairly than any judge, and allocates resources more efficiently than any elected body, why preserve human deliberation?
The Talmud’s answer to that temptation remains radical. In a celebrated legal debate, Rabbi Eliezer called on a heavenly voice, which declared that “Rabbi Eliezer is right in every instance.” You’d have thought that would settle things. Not so. Rabbi Yehoshua stood and pronounced: “It is not in heaven!” Human deliberation, he insisted, has value independent of whether it produces correct outcomes. The Talmud reports that “God smiled and said: My children have defeated Me.”
The implication is not that superior tools should be rejected. A tradition that values healing need not reject better diagnostic tools. But it would be far more reluctant to outsource self-government, moral judgment, and the weighing of collective priorities. A civilization that outsources moral reasoning to a superior intelligence, even a correct one, has surrendered something that better outcomes cannot replace.
The ancients also warned that the journey toward a better age would be fraught. The Jewish tradition called this passage hevlei Mashiach, “the birth pangs of the Messianic age”: “Insolence will increase... prices will rise... the government will turn corrupt... truth will be absent... the young will shame the old.” We may be entering the early stages now.
But Hebrew prophecy differs from Greek oracles. The Greek oracle is vindicated when its prediction comes true. The Hebrew prophets’ dark visions were warnings rather than predictions, meant to spur action, not resignation. In that sense, when a biblical prophecy came to pass, the prophet had failed his mission.
So the Jewish foretelling of hevlei Mashiach is not inevitable. The subtext is preemption. Mechanisms to share abundance widely; education that values learning independent of jobs; unpaid frameworks for productivity, creativity, and public service; and domains where human judgment remains supreme. These can form a birth plan for the coming era.
Thousands of years before AI, the ancients imagined much of it – self-operating tools, autonomous weapons, answer-giving superintelligence. They sat with the harder questions, too. Who abundance is for, what it is for, and whether human judgment can be delegated to something greater than us.
As AI reshapes everything, it is steadying to know we are not starting from scratch. The ancients knew nothing of our technology. But our most pressing questions aren’t technological. They are questions of meaning and judgment, of how to live and what to live for.
On those, they’ve bequeathed us something timeless. And timely.
The writer is the CEO and co-founder of Lemonade (NYSE: LMND) and chairman of the MOSAIC AI Policy Institute (mosaic.org.il).