There is much talk today about the safety of artificial intelligence. Underlying that concern is a deeper question: Can we achieve forms of intelligence, whether artificial or natural, that do not simply reproduce the compulsion of our civilization to use knowledge for control and domination—of the natural world or of one group by another? The very concept of intelligence underwriting our technological civilization may be unsafe because it is one-sided, linear, and brittle.
An implicit aim of modern science has been to understand the world completely, as though nature were a machine derived from a set of blueprints. In principle, this project promises to reduce natural reality to well-defined systems that can be encoded algorithmically. Once a system has been understood in these terms, technology can be used to manipulate it, reproduce its structure and functions, and ultimately bring it under human control. The desire to understand and reproduce nature through technology, however, has a broader context. It is part of an ancient aspiration to establish the conscious self as something separate from nature and sovereign over it—even immortal. The attempt to create artificial life and artificial mind can be understood not only as a technological project but also as a kind of spiritual quest: an effort to stand apart from nature, transcend its limits, and appropriate its powers. In other words, to be godlike in power, if not in virtue. The same mentality underlies the capitalist project to appropriate natural “resources” for profit.
Organisms strive to maintain their internal organization in the face of external changes. It is therefore natural to conceive of the world as an affordance, a field upon which an organism acts in order to sustain itself. Humans push this orientation to an extreme, however, so that it becomes the conscious self’s attempt to be master of its destiny. Everything outside the ego—including the body itself—is then reduced to an object to be used and controlled. Freedom becomes identified with the detached mind and its capacity to manipulate, while limitation is identified with whatever can be manipulated: the material world, nature, the bodies and minds of others. This is a distorted version of homeostasis: instead of allowing oneself to change in response to the world, one attempts to transform the world while remaining unchanged.
A radically different approach is a receptive stance. Rather than treating the world, and other people, primarily as objects to be controlled, there is willingness to be changed by what is encountered. The world then becomes not simply a field of obstacles to be overcome or resources to be exploited, but a field of possibilities for growth. Such receptivity is not passive. It is a different version of agency, in which intelligence consists not in imposing a predetermined order upon reality but in remaining responsive to a reality that continually exceeds our ideas about it.
The logic of control tends toward linear thinking, and one of the fundamental weaknesses of linear thinking is its vulnerability to apparently minor errors or deviations. In a linear system, causes and effects remain roughly proportionate: a small disturbance produces a small consequence. In complex systems, however, a small disturbance that is initially reversible can propagate through a network and become amplified into a large-scale disaster.
Robustness is the capacity of a system to absorb deviations while maintaining its essential organization. In living systems, this means actively preserving and reproducing a form despite continual changes in inputs and environment. This biological capacity— autopoiesis—is the self-production and self-maintenance through which a living system continually constitutes itself and its robustness. A counterpart in ordinary human life might be called common sense. Common sense is not a fixed collection of rules. It is the capacity to remain oriented amid uncertainty and chaos. Specifically, it includes the ability to recognize when a strategy or procedure is producing an absurd result, and to adjust behavior accordingly.
This raises a fundamental question for artificial intelligence and technology: To what extent can robustness, or common sense, be incorporated into programs that are not themselves genuine autopoietic agents—artificial organisms? A machine can certainly be designed to resist errors and hijacking, adapt to changing inputs, and preserve goals specified by its designers or users. Yet unless its programming can continually update itself in relation to the real world, its robustness will always be limited. In living systems, robustness is not simply an additional feature that can be inserted into an otherwise brittle mechanical structure. It arises from the very organization of life.
An organism cannot be disassembled, conceptually or materially, into a list of properties or parts, and then reconstructed by assembling those into an artificial version. The aspiration to create artificial organisms or brains depends, among other things, on the assumption that natural organisms can first be understood as mechanisms. But mechanism is a way of looking: a powerful method of abstraction, with technological uses but also definite limits. Machines are products of human design and definition, reflecting the human desire to control. Although machines can be made from durable materials and engineered to tolerate particular disturbances, they are not robust in the general and active way that organisms are. An organism is not a product of human design, nor simply the sum of its identifiable properties. Its robustness is self-creating and inseparable from its living embodiment, its history, and its ongoing interactions with the world. Its intelligence cannot be isolated from that history, and its agency cannot be reduced to a collection of algorithms.
A genuinely robust artificial intelligence, if such a thing is possible, would therefore not be one that perfectly follows rules or maximizes a predetermined objective. Rather, it would have the ability to recognize when its programming fails to correspond to reality. In short, it would have common sense: the ability to remain responsive to changing reality rather than becoming trapped within its own models. This is closely related to the receptive stance toward nature described above: an intelligence capable of revising itself in response to what it encounters, rather than rigidly imposing its categories upon the world and being entrapped by them.
The political implication is significant: society should be robust. If technology represents humanity’s attempt to become the author of its own fate, then genuine wisdom requires recognizing the limits of that project. A technological civilization capable of surviving its own power would need moral development along with technical development. Only a society capable of self-reflection, self-restraint, cohesion, and long-term responsibility for the future could safely pass through a technological phase without destroying the conditions for its own existence. Technological progress without the brake of common sense magnifies the consequences of error. It increases power without necessarily increasing wisdom. The more powerful and complex the system, the more catastrophic even a small failure can become.
The alternative to technological expansion is not stagnation, however, but a robust equilibrium. Rather than rushing toward every new frontier—whether in space, artificial intelligence, biotechnology, or economic growth—we might pause to recover forms of social and ecological balance without which technological civilization is unstable. This would also require us to reconsider what we mean by freedom and by growth. If freedom is defined exclusively as the power of individuals to impose their will, then the destruction of nature and society is guaranteed. But if freedom means the capacity to participate with good will in relationships of mutual dependence, then restraint can become a form of freedom: the freedom to act for the benefit of a larger whole. Growth can be qualitative.
Agency, in this view, does not have to mean domination. Organisms do not survive by eliminating all deviation or controlling every aspect of their environment. They survive through their capacity to respond to change without losing themselves, to incorporate disturbances, and to remain organized in the midst of chaos and uncertainty. A mature intelligence might therefore understand agency not as absolute control but as the ability to participate effectively in a world beyond itself.
Perhaps the wisdom that technological civilization needs is to recognize that a world made entirely in our own image would not necessarily be a better world. The challenge now is not to be masters of nature, by appropriating it to our purposes and recreating it, but to become capable of living in relation to something that is neither our creation nor our possession. Only then might the human project become more than an elaborate attempt to escape limitation. Intelligence, then, would not be mastery but living well within a world we can never completely control.