Praise Against the Machine

What Cantors Can Teach Us About the Dangers of A.I.

Photo by Marko Horvat on Unsplash

I have a saying about AI. Here it is:

Whatever we give to AI, we will forget how to do.

Whenever a new technology takes responsibility for an essential task, there’s a strong chance that human beings will come to depend on that technology.

Take driving: In the twentieth century, drivers read maps and developed an internal compass for navigating the road. But this once-internal learning has given way to a dependence on satellite navigation. If the GPS breaks down or our phone dies, many of us become lost and utterly unable to get our bearings. Brainpower has been replaced by machine power.

This is true about countless other skills that our modern era has made easier or obsolete through technology, like ironing (dry cleaners), developing film (digital cameras), sewing (cheap clothing), hunting (supermarket), building a fire (electric lights/fire starters), driving a manual (automatic transmission), and car repair (AAA). We may retain some old muscle memory for these things, but our abilities will inevitably atrophy from lack of use.

While these innovations represent the loss of technical skills, in the twenty-first century we face an even graver threat: the loss of our intellectual, emotional, and spiritual capacities. Over the last decade, the addictive, social media-enhanced smartphone has decimated the social skill, attention span, learning capacity, and mental health of younger generations.

Yet AI represents an even more significant challenge. While promising the brightest heaven of invention, it threatens to obviate the most fundamental of human abilities: the ability to think.

As a cantorial historian, I have followed technological changes and their long-term effects on human behavior in the realm of prayer. And a look back at the story of how cantors became prayer specialists reveals an eerie echo of the very dangers we face today.

There is perhaps no cognitive feat more uniquely human that the act of prayer. Addressing the transcendent, we place our individual human longings within a vision of faith and a community of spiritual practice. For a millennium of rabbinic Judaism, prayer was largely an orally-transmitted form, held in the bardic memories of Jewish leaders and laity alike and embellished over time by cantor-poets and scholars. The congregation, having memorized prayers from a very young age, relied on the specialization of the cantor (often a rabbi as well) for religious poetry and traditional song. This was also an age of handwritten manuscripts, in which the cantor was likely the only one with a prayerbook in hand.

In this environment, technological innovation changed the mental processes of prayer. As Sefer Lev Tov (Prague 1620) testifies, the advent of cheap, printed siddurim dissolved the traditional division of liturgical labor as everyone began to sing aloud. After all, everyone now had a prayerbook, making them look just like a medieval cantor. This was a tech revolution that democratized leadership, not hoarded it.

But at the same time prayer became a more technically-specialized enterprise, requiring the cantort to contemplate many new layers of mystical intentions (kavannot) during the service. At the same time, musical complexity was growing in the synagogue, offering service leaders new specialized music to beautify the service. As prayer became professionalized in early modern Europe, cantors took on new musical abilities and assistants, specialized (and even esoteric) knowledge, and accreditation from trusted practitioners. As a result, new heights of artistic achievement were reached and great beauty was released into the world.

But this beauty came with a dark side. As the role of prayer leader became more and more distinct from that of the congregation, the act of prayer began to appear more and more as the province of the clergy alone.

A.Z. Idelsohn with informants from his ethnographic journeys in Mandatory Palestine (Source: National Library of Israel (MUS 0004 K 07_1).

The Jewish musicologist A.Z. Idelsohn was shocked to discover this during his ethnographic work among the Jews of British Mandate Palestine. Encountering lay-driven communities with democratized knowledge of the traditional prayers, Idelsohn expressed shock by the contrast with the professionally-led communities of Western Europe:

“Thus you have almost no Yemenite, Moroccan or Iraqi, Lithuanian or Pole who is not versed in the prayers and traditional melodies and is qualified to go before the ark, at the same time that almost no Jew in Germany, France, Italy, and even amongst European Sephardim that are able to go before the ark. For this reason, in Yemen, Iraq, Morocco, and Lithuania the role of prayer [leader] and sheliach tzibbur remains in its ancient form, as it was handed down from Moses, Our Teacher, a matter that every person of Israel should be fitting to do, the property of all and a folk-based, democratic inheritance; at the same time that Germany, Spain, Italy, and the like have almost lost the memory of the lay prayer leader, and thus make the cantor the only prayer leader, on behalf all of his community, and without him one would find neither hand nor foot in the synagogue.”¹

This atrophy of prayer skills observed by Idelsohn among Western European Jews continued into our own era, this time with masses of Eastern European Jews on American shores. Cantor Samuel Rosenbaum, who presided over the Cantors Assembly for the second-half of the twentieth century, wrote regularly against the slow atrophy of lay competency in traditional prayer:

“More and more we miss the reassuring hum of a congregation davening; the comforting continuum of the davening murmerei. We hear, less and less, the natural, unrestrained praying sounds which daveners create as they bite into a new tefillah with a will. The silence which now follows the conclusion of the Cantor's chant, or of a choral piece, grows more deafening and disturbing.”²

Can we cantors really be blamed for such a development? The erosion of traditional prayer in many non-orthodox (and even some orthodox) communities is a long-lamented phenomenon, relating to glacier-sized changes in Jewish observance, faith, Jewish education, identity, and the rhythms of modern life. The decline of synagogue attendance and traditional prayer knowledge is a symptom of these larger issues, one in which the cantor plays but one part.

Yet what is notable is that traditional Jewish communities, whether egalitarian or orthodox, have managed to convey to their followers that prayer leadership is not a professional matter — it is a Jewish matter. Prayer requires high standards and religious presence to lead. Yet it is not in the academic heavens, nor in the cantorial sea, but a spiritual practice that is near to the common person’s heart.

A number of cantors have succeeded in conveying this to their congregation while still holding specialized roles in music and prayer. I grew up in a synagogue like this, in which prayer leadership opportunities were shared with the permanent hazzan, and in which bar/bat mitzvah children led as much of the service as possible. Those skills prepared me not just for my eventual profession, but for my sense of individual agency as a Jew in relationship with God.

Looking back at the long arc of cantorial history, from Sefer Lev Tov to Idelsohn to Rosenbaum, I am convinced that specialization is both dazzling and dangerous. It unleashes the artistic and creative abilities of a devoted professional, and yet one must work against its solipsistic temptations in order to inculcate the common, democratic inheritance of prayer among all of God’s children.

---

Artificial Intelligence is the ultimate specialist. It is dazzling—and dangerous—on a completely astronomical scale.

Just as the professionalized cantorate ran the risk of communicating that non-professional Jews no longer needed to pray, A.I. runs the risk of robotically professionalizing countless aspects of human life, with the unintended consequence of decimating those abilities in the human population. Just see the damage that has already been done:

• By asking AI for help with schoolwork, our educational system is at risk of collapse.

• By asking AI to think for us, human ability to do critical thinking is tanking.

• By asking AI to create the arts — audiobooks, music, literature, and even film — we risk eliminating these creative positions from our economies.

• By asking for AI to guide our spiritual lives, our individual capacity for spirituality is weakening leading to AI-based religious cults.

AI advocates like Tyler Cowen describe a future over the next two decades where these are regular realities to which we simply adapt. And like the cantorate’s specialized musical ingenuity, I don’t think that AI is going away, and believe that both have a lot to offer the future of humanity.

But I do think that the trajectory that we are on looks like a slow march into a voluntary version of the Matrix. For older generations whose critical thinking was formed in a pre-internet era, AI is an additive tool. But for younger generations, it is quickly becoming a “permanent companion,” outsourcing their critical thinking, emotional needs, and spiritual longings to a soulless machine.

In this way, perhaps we should treat AI like that old adage about kabbalah— something that you should not study until you are forty, or until one has developed intellectual and spiritual maturity.

At a minimum, we should take a lead from this dark and dazzling history of prayer. Like a good synagogue, we should invest in the critical thinking, emotional depth, and spiritual capacity of our young people. We want them to think for themselves, connect with curiosity and empathy with other human beings, and reach out to the Creator in whose image they are fashioned.

These are things that we should never invite AI to help us forget.

1. Abraham Zvi Idelsohn,“HaHazzan BeYisrael,” HaToren, Vol. 10 no. 3 (1923): 62.

2. From “Towards a New Vision of Hazzanut,” 1989. Published in Tefillat Shmuel: Selected Writings of Hazzan Samuel Rosenbaum. Ed. Matthew Austerklein. Cantors Assembly: 2018).