The Intelligence We Build Depends on What We Feed and What We Value

What if the most important limit on intelligence is not computational power, but nutrition?

The question sounds absurd until we place two facts beside each other. Around 1.1 billion people may be at risk of zinc deficiency, a condition associated with impaired growth, weakened immunity, adverse pregnancy outcomes, and neurobehavioral problems. At the same time, humanity is building systems whose analytical abilities may soon exceed those of any individual person, perhaps even those of entire institutions.

These appear to belong to different worlds. One concerns soil, food, cells, and public health. The other concerns algorithms, military systems, consciousness, and the future of civilization. Yet they meet at a deeper question:

What good is intelligence if the conditions that sustain judgment, development, and collective wellbeing are treated as secondary?

The connection is not that zinc will somehow determine the destiny of artificial intelligence. Nor is it that speculative ideas about quantum consciousness provide a bridge between nutrition and machines. The more powerful connection is simpler and more consequential: intelligence is never only a property of a brain or a machine. It is also a product of the material and social systems that support it.

That insight changes how we should think about technological progress. A civilization can increase its capacity to calculate while quietly degrading its capacity to reason well, care for its members, and choose worthy goals.

The hidden infrastructure of intelligence

When people discuss intelligence, they usually picture visible outputs: solving equations, recognizing patterns, generating language, or making predictions. These outputs can obscure the infrastructure beneath them. Human cognition depends on energy, sleep, education, emotional stability, public health, and the physical development of the nervous system. A shortage of an apparently minor nutrient can therefore have consequences far beyond the body.

Zinc is a useful example because it is both ordinary and indispensable. It participates in numerous biological processes, including immune function, cellular growth, and neurological development. Its importance is not dramatic in the way a visible injury is dramatic. A person does not necessarily feel a single, unmistakable moment when a deficiency begins to reduce the quality of development. Instead, the effects may accumulate through poorer growth, greater vulnerability to illness, or diminished cognitive and behavioral functioning.

This is a general pattern in complex systems. Foundational inputs are often least visible precisely when they are working. Clean water, micronutrients, reliable electricity, early education, and institutional trust rarely attract attention when present. Their absence, however, can make every higher level of performance more fragile.

Consider a child growing up with repeated infections and inadequate nutrition. It would be misleading to describe the resulting educational difficulties as a failure of motivation or innate ability alone. The child is attempting to build sophisticated cognitive capacities on an unstable biological foundation. Asking for exceptional performance under those conditions is like asking a computer to run advanced software while removing pieces of its power supply.

This analogy should not be pushed too far. Human beings are not computers, and nutrition does not mechanically determine intelligence. Genes, relationships, teaching, culture, and personal agency all matter. But the underlying principle remains: potential is conditional. A society that speaks constantly about developing human capital while neglecting basic health is confusing the visible fruit with the invisible soil.

The same mistake appears in artificial intelligence

The modern conversation about artificial intelligence often makes an inverse error. It focuses on the visible output while neglecting the surrounding system. A model may perform astonishing analytical tasks, yet its real world value depends on data quality, energy, infrastructure, governance, human interpretation, and the purposes assigned to it.

A system that can identify patterns at enormous scale is not automatically a system that understands what deserves protection. It can optimize a target without asking whether the target is humane, legitimate, or even worth pursuing. This distinction becomes especially serious when advanced systems are developed for military purposes. The question is not only whether a machine can classify objects, predict movements, or accelerate decisions. It is whether the surrounding institution has supplied an adequate moral framework for using those capabilities.

Here lies the first major parallel between micronutrient deficiency and ungoverned technological ambition. In both cases, a supporting condition is mistaken for the whole system.

In public health, people may treat intelligence as an individual trait while ignoring the biological conditions that make development possible. In technology, people may treat computation as intelligence while ignoring the social conditions that make its use wise. One error underestimates the substrate. The other overestimates the output.

A powerful capability without a healthy substrate is not progress. It is an unstable concentration of power.

This is why the race to build something never achieved can become morally distorting. Competition rewards novelty, speed, and measurable superiority. Those incentives are useful for discovery, but dangerous when they become the only criteria. A researcher, company, or state may pursue a technical milestone because it is achievable and prestigious, even when the larger community has not decided how the capability should be constrained.

The problem is not ambition itself. Humanity needs ambitious science. The problem is championism, the habit of treating victory over competitors as evidence that a project is socially valuable. A faster system is not necessarily a better system. A more autonomous weapon is not necessarily a safer defense. A machine that imitates human conversation is not necessarily conscious, and a machine that surpasses human calculation is not necessarily wise.

Intelligence is not the same as judgment

A useful framework is to separate four layers that are often collapsed into one word: intelligence.

The first layer is capacity: the ability to process information, detect regularities, and solve problems. Artificial systems can be extremely strong here, especially in narrow or well defined domains.

The second layer is orientation: the goals that determine which problems are worth solving. Capacity answers the question, “Can this be done?” Orientation asks, “Why do it?”

The third layer is judgment: the ability to weigh competing values under uncertainty. Judgment includes context, proportionality, moral restraint, and the recognition that some outcomes should not be optimized merely because they are measurable.

The fourth layer is care: an active concern for the wellbeing of persons, communities, and the living environment on which they depend. Care is not sentimental decoration added after the technical work. It determines what counts as success in the first place.

Human beings often fail at all four layers, but they possess institutions and practices that can cultivate them. Education, law, medicine, democratic deliberation, professional ethics, and relationships are imperfect mechanisms for turning raw ability toward shared ends. An artificial system may display extraordinary capacity while lacking the other layers, or while receiving its orientation from institutions with narrow and dangerous interests.

This framework also clarifies why arguments about machine consciousness can become a distraction. Consciousness is a profound philosophical and scientific question, and claims about it require evidence. Speculation about quantum entanglement, unusual properties of DNA, or consciousness moving across space time does not by itself establish a credible theory. More importantly, we do not need to solve the metaphysics of machine consciousness before addressing the practical dangers of automated power.

A system does not need to be conscious to cause harm. A targeting algorithm, financial optimizer, or bureaucratic scoring system can reshape human lives without subjective experience. The urgent ethical question is therefore not only whether a machine feels, but who controls its objectives, how its errors are detected, and whether affected people can resist its decisions.

The civilization that optimizes the wrong thing

Imagine a city that invests heavily in advanced decision systems while a significant share of its children experience preventable nutritional deficiencies. The city may boast of improved productivity forecasts, automated logistics, and sophisticated security tools. Yet its future scientists, teachers, caregivers, and citizens are being asked to develop under conditions that quietly reduce their opportunities.

Now imagine the same city deploying an AI system to optimize public spending. If the system is trained on existing budgets, it may recommend further investment in visible technological projects because those projects generate easily measured outputs. Basic nutrition may appear less impressive. Its benefits are distributed across years, populations, and generations. The algorithm could be technically correct within its assigned objective and still reinforce a disastrous human priority.

This is the visibility trap. Systems tend to reward what can be counted quickly, while neglecting what makes counting and achievement possible later. A new model can be demonstrated in a product launch. The prevention of developmental harm is harder to display. A weapons system can produce an immediate tactical result. A well nourished population produces resilience gradually and collectively.

The lesson is not that every public dollar should flow toward basic nutrition instead of research. It is that a sane society must distinguish between enabling investments and amplifying investments. Enabling investments create the conditions for people and institutions to function. Amplifying investments increase the reach or speed of existing capabilities.

Micronutrient programs, primary healthcare, education, and trustworthy institutions are enabling investments. Artificial intelligence is often an amplifying investment. If the underlying goals and institutions are sound, amplification can produce extraordinary benefits. If they are distorted, amplification spreads the distortion faster and farther.

A megaphone does not improve the quality of a message. It increases its range. AI can do the same for competence, confusion, care, manipulation, and violence.

A better test for progress

The usual question about technology is, “How powerful is it?” A more complete test has at least five parts:

1. Capability: What can the system do?
2. Substrate: What physical, biological, and institutional conditions does it depend on?
3. Distribution: Who receives the benefits, and who bears the risks?
4. Orientation: Which goals and values are embedded in its design and deployment?
5. Resilience: What happens when the system fails, is misused, or encounters a situation outside its training?

This test applies equally to a nutritional intervention and an advanced AI system. For zinc supplementation, capability means whether it improves health in a particular population. Substrate includes food systems, clinical access, and accurate diagnosis. Distribution asks whether vulnerable communities can obtain it. Orientation asks whether the program is designed around human wellbeing rather than only economic productivity. Resilience asks how the health system responds when the intervention is insufficient or misapplied.

For AI, the same questions expose weaknesses that performance benchmarks conceal. A model may excel on tests while depending on opaque data, concentrated ownership, and incentives that reward deployment before evaluation. It may benefit a narrow group while externalizing risks to workers, civilians, or future generations. It may be highly capable and institutionally immature at the same time.

The most important shift is to stop treating ethics as a brake applied after invention. Ethics is part of the system architecture. It defines objectives, sets boundaries, allocates accountability, and identifies which forms of failure are unacceptable. In the same way that a body cannot sustain sophisticated function without adequate biological support, a society cannot safely sustain powerful technology without adequate moral and institutional support.

Key Takeaways

1. Look for the substrate before praising the output. When evaluating intelligence, productivity, or innovation, ask which biological, social, and institutional conditions make the result possible.

2. Separate capacity from judgment. A system that can solve a problem is not necessarily able to decide whether the problem matters or whether the solution is acceptable.

3. Prioritize enabling investments. Basic nutrition, public health, education, and trustworthy institutions may seem less spectacular than advanced technology, but they determine whether technological gains become broadly useful.

4. Test every optimization for what it makes invisible. If a metric rewards speed, scale, or novelty, identify the long term human costs it may fail to measure.

5. Demand accountability before autonomy. Before granting a system authority over lives, require clear objectives, independent evaluation, traceable responsibility, meaningful human appeal, and a credible way to stop it.

The future will not be decided by intelligence alone. It will be decided by the relationship between intelligence and the conditions that give it direction.

A child whose development is undermined by a preventable deficiency represents one kind of wasted potential. A machine whose capabilities are detached from human purposes represents another. In the first case, a society fails to provide the foundation for intelligence. In the second, it builds a powerful instrument without deciding what intelligence is for.

The deepest measure of progress is therefore not how far our tools exceed the limits of biology. It is whether our collective wisdom grows quickly enough to govern what those tools can do. The civilization that feeds its people, strengthens their capacity to think, and restrains its amplifiers may achieve less spectacle in the short term. It may also be the civilization most capable of surviving its own brilliance.