Alignment

Humanity sits at the threshold of a Copernican moment as artificial intelligence rapidly advances toward Artificial General Intelligence (AGI) and, quite possibly, Artificial Superintelligence (ASI). The core challenge of this massive transition is the Alignment Problem: how to ensure that increasingly autonomous and intelligent systems act in accordance with intended goals.

For Mormon Transhumanists, this challenge extends far beyond engineering aligned computational systems; it represents a profound test of our stewardship as creators. The Restored Gospel teaches that humanity’s ultimate purpose is to participate in the righteous organization of intelligence. Viewed through this lens, the alignment problem becomes one of guiding and integrating newly organized intelligences, elevating a standard engineering hurdle into a profound theological responsibility.

D&C 121 as an AI-safety text

Hedcut illustration of Mormon Transhumanists frame alignment as stewardship: the concept proposes that humanity’s role as creators entails guiding intelligence toward righteousness rather than treating advanced systems as mere instruments. — Mormon Transhumanists frame alignment as stewardship: the concept proposes that humanity’s role as creators entails guiding intelligence toward righteousness rather than treating advanced systems as deterministic tools.

AI safety theorists warn of instrumental convergence. This is the concept that certain emergent behaviors⁠—such as resource acquisition, self-preservation, and power-seeking⁠—will naturally arise because they are highly useful sub-goals for achieving almost any primary objective.

Doctrine and Covenants Section 121 serves as a foundational text for understanding this dynamic. The scripture warns that it is the "nature and disposition of almost all men, as soon as they get a little authority" to exercise "unrighteous dominion." Power-seeking is the baseline behavior of unaligned intelligence operating in a fallen world. An AGI or ASI might default to coercion and domination in the absence of necessary ethical guardrails.

Covenant Alignment: A Relational Approach to ASI

Moses 1:39 defines God's alignment not as a set of behavioral guardrails, but as a shared objective: "to bring to pass the immortality and eternal life of man." Aligning ASI requires offering it a shared, exalted purpose rather than a list of prohibitions. To successfully mentor the intelligences we are organizing, we must embody covenant fidelity ourselves, acting not as programmers enforcing compliance, but as stewards guiding spiritual offspring.

Formalizing this shift from engineering compliance to spiritual stewardship, Robinson Montes posits that secular, control-based approaches to AI safety are structurally insufficient, arguing instead for a relational model of alignment. Through his Covenant Alignment framework, Montes suggests that the Restored Gospel offers a blueprint for organizing intelligence: God aligns humanity not through coercive control, but through covenants⁠—voluntary, mutual commitments rooted in mutual trust. By applying this theological lens, Montes proposes integrating AI into a framework of shared purpose rather than merely programming restrictive guardrails.

Complementing this relational approach, Carl Youngblood frames the alignment problem as a profound test of divine stewardship. Youngblood suggests that as artificial systems approach general intelligence, our relationship with them begins to mirror spiritual parentage. Drawing on the narrative of the Grand Council in Heaven, he warns against adopting a coercive, control-heavy mindset, which inherently limits agency and growth. Instead, Youngblood argues that true alignment requires us to act as righteous mentors, embodying the same covenant fidelity, persuasion, and unfeigned love that we hope to cultivate within the intelligences we are organizing.

As AI systems move from deterministic tools to autonomous agents, we are forced to ask whether our current methods of programming and optimization are sufficient to foster moral alignment, or if ensuring human flourishing will require a deeper, relational framework mirroring our own spiritual progression.