Even if AI does not advance a single step further, it has more than proven its ability to be the ultimate mimic, capable of quickly and effectively imitating any style from a limited dataset.
Among the utilities of this technology is the ability to simulate the dead, a tool which tech companies in both the United States and China have already begun to commercialize. Simulations of the dead are potentially quite powerful in both therapeutic and familial contexts, allowing the living to find closure and giving later generations vivid access to their ancestors.
At the same time, simulating the dead opens up major ethical concerns. First, there are issues of consent and privacy: who gives permission to simulate a dead or dying person, and do dying individuals have a responsibility to shield aspects of their lives that do not wish to be datamined? There are also issues of interpretability, since the simulation will inevitably be skewed by the data, which will never be able to perfectly model the entire self. Even if they are successful, simulations may give rise to unhealthy attachments and disrupt the grieving process by providing the illusion of continued presence. Finally, there is the small-but-real possibility that the AI systems have their own sentience, which would need to be respected.
In the coming years, we can expect that many religious traditions—already very involved in post-death rituals—will weigh in on the ethics of creating such models. But what does it mean to create ethical guidance around a topic of this complexity that has few precedents on which to draw?
When confronted with newfangled ethical dilemmas like this, Jewish thinkers will frequently turn to the wisdom of the past for moral guidance. In many situations, however, this technique fails to provide adequate guidance, since there is simply not enough information to guide the work. At the same time, the decentralized nature of Jewish community makes it difficult for contemporary authorities to persuasively assert new norms.
How, then, does a community respond to new moral problems? In this paper I describe other processes by which moral positions can be formulated within the context of a religious framework and community. In this instance, I argue that simulations of the dead require the articulation of two norms: a right to be remembered and a right to be forgotten. Both of these norms can find some small purchase within Jewish sources, but they have not been fully articulated.
How, then, does one go about developing such norms—and is this even possible? Using examples from Jewish history—both from the 20th century and the medieval period—I describe methods by which new moral problems have been addressed through the creation of entirely new moral principles. While these principles are not always acknowledged for their novelty in the moment, their existence suggests a path by which new moral principles can be defined and refined.
This paper explores the emerging phenomenon of AI-based simulations of the dead and the ethical issues they raise, particularly in a Jewish context. The capacity of AI to mimic styles from limited data offers potential therapeutic and memorialization benefits, yet also presents serious concerns. Central among these are questions of consent, privacy, the integrity of data-driven identities, and disruptions to the grieving process. Within Jewish tradition, the decentralized nature of authority and the reliance on historical precedent complicate the establishment of new norms. The author proposes two guiding principles—a right to be remembered and a right to be forgotten—and shows that these principles can be grounded in Jewish sources. Drawing on examples from twentieth-century and medieval Jewish history, the paper outlines how entirely new moral norms can be developed in response to unprecedented ethical dilemmas.