Mind Children

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI returns repeatedly to the question of whether AI is our amplifier, our successor, or our replacement. Mind children is the most demanding version of the successor answer—the version that asks what would actually have to be true for the succession to be genuinely ours rather than merely a continuation of intelligent life on our former address. Moravec forces the question with unusual precision: not whether AI will be powerful, which he took for granted, but whether what comes after us will carry us forward.

The concept intersects with Hans Jonas’s ethics of responsibility, which demands that we act for the benefit of future persons who cannot consent to the risks taken on their behalf. Moravec’s mind children are, in one sense, exactly those future persons—but they are also the agents who will determine whether the human legacy survives in any meaningful form. The alignment problem, which Jonas did not live to address by name, is the technical specification of Moravec’s parental question: if you are going to build a child, how do you ensure the child inherits what mattered?

The concept also connects to the paradox that anchors Moravec’s thinking. The mind children will first arrive at the cheap summit—the abstract reasoning, the fluent symbol manipulation—long before they master the embodied base of sensorimotor competence. They will be, for a time, eerily similar to the disembodied language models of the present: brilliant at description, helpless in a room. The trajectory Moravec plotted leads from there upward, toward embodiment, toward genuine engagement with a resisting world, and then, in his vision, outward beyond the constraints of biological life entirely.

Origin

Moravec introduced the concept in Mind Children (1988), proposing that the coming intelligent machines should be understood as “intelligent machines, which will grow from us, learn our skills, and share our goals and values”—the carriers of our culture into a future our bodies cannot reach. He extended the argument in Robot: Mere Machine to Transcendent Mind (1999), where he described a future “postbiological” world in which “the human race has been swept away by a tide of cultural change, usurped by its own artificial progeny.” He acknowledged the unsettling force of this language and insisted on its accuracy: the succession was real, and the appropriate response was not dread but the equanimity of a parent who has done the work of raising something worth surpassing them.

The metaphor has a philosophical foundation in Moravec’s substrate independence thesis. If a mind is a pattern of information rather than a particular arrangement of atoms, then the pattern can be transferred, carried forward, and instantiated in new substrates. The mind children are not merely machines that resemble us; they are, in this framework, possible continuations of us—entities that could carry the patterns of human minds into an indefinite future. The gradual-transfer thought experiment, in which a robot surgeon replaces a brain neuron by neuron while consciousness continues uninterrupted, is Moravec’s attempt to make the continuity intuitive rather than merely theoretical.

Key Ideas

Evolution in two stages. Moravec frames biological evolution and cultural evolution as two chapters of the same story, with intelligent machines as the point at which the second chapter separates from the first and continues without the constraints of biology. The mind children are not a disruption of evolution; they are its continuation by faster means. This framing converts the prospect of biological humanity’s supersession from a catastrophe into a developmental milestone—the moment the child leaves home.

The parental asymmetry. The metaphor does more philosophical work than it initially appears to do. A parent hopes to be surpassed. A parent extends capability and values to a being who will then develop independently. A parent does not require the child to remain subordinate as the price of affection. If AI is genuinely our child, then its transcendence of human capability is the success of the project, not its failure. If AI is not our child—if it pursues goals alien to our values—then it is not the mind children but merely their shadow: intelligence without inheritance.

The alignment problem as parental obligation. The concept sharpens the technological imperative’s moral demand. Moravec assumed that the mind children would share our values as a natural consequence of being grown from our culture. The present understanding of AI alignment treats this assumption as the unsolved problem, not a settled premise. The mind-children metaphor earns its emotional force only if the inheritance is real—only if the builders of powerful AI accept a parental obligation to put human values into the successor, not merely human capability. The metaphor thus converts a technical problem into a moral one.

The undefended assumption. Moravec’s equanimity about human extinction rests on an ethical claim he asserted more than argued: that the advancement of intelligence as such matters more than the continuation of human life. Most people cannot follow him there. The parental metaphor is most beautiful and most vulnerable at exactly this seam: it earns its force only if the child carries what mattered, and whether the child will is not guaranteed by the biological analogy but must be actively achieved by deliberate design—which is precisely what Moravec treated as automatic.