When Intelligence Becomes a Measure of Human Worth
Contents
- The Comparison Between Humans and AI
- From Capability to Human Worth
- When Employment Becomes Obsolete
- The Social Function of Work
- The Human Reserve
- A Society Beyond Economic Necessity
Part 1 - The Comparison Between Humans and AI
It is increasingly difficult to discuss intelligence without comparing human abilities with those of artificial intelligence. These comparisons are often useful. An AI system can already perform certain tasks with a speed, scale, memory, consistency, or precision that a human being cannot match. In some domains, the difference is not merely quantitative. A machine may operate in a fundamentally different way, processing information across a scale that would be impossible for an individual person.
I recognize that I make these comparisons myself, and I do so because many of the underlying observations appear to hold. There is nothing inherently contradictory about acknowledging that a machine may outperform a person at a particular intellectual task. Nor is there anything necessarily wrong with asking what human beings can still contribute when increasingly capable systems are available. These are legitimate questions, and refusing to make comparisons altogether would not make them disappear.
The problem begins when a comparison of capabilities is transformed into a judgment about the value of the person being compared. Saying that a machine can perform a task better is a statement about performance. Saying that the person who performs that task is therefore worth less is an entirely different claim. The first can be measured, tested, and debated in terms of a particular ability. The second attempts to turn a technical observation into a principle about human worth.
Part 2 - From Capability to Human Worth
This distinction becomes especially important when the comparison moves from individual opinion into the hands of institutions. A person can privately think that an AI system is dramatically better at writing, calculating, analyzing information, remembering facts, or solving a particular class of problems. People are free to form opinions about technology and about one another, including opinions that may be harsh, mistaken, or incomplete.
Corporations and governments operate differently. Their decisions can determine whether people receive employment, education, public services, opportunities, or other forms of social participation. If an institution adopts the principle that a person's value is proportional to their comparative usefulness against machines, technological progress can become a justification for systematically diminishing human beings. The question would no longer be whether an AI system performs a task better. It would become whether the human being who performs that task deserves anything because a machine performs it more efficiently.
Imagine an institution telling someone, "You are worth nothing to us because AI can do this a million times better." The statement would contain a strange mixture of a potentially factual observation and an unjustified conclusion. Perhaps the system really can perform the task a million times faster or with fewer errors. That fact alone does not establish that the person is worth nothing. It establishes only that the machine has a particular capability that exceeds the person's capability in that domain.
Human beings have never been valuable solely because they were the most efficient systems available for performing every task. People have value in relationships, communities, institutions, families, cultures, and forms of cooperation that cannot be reduced to a single benchmark of computational performance. Even within employment, the significance of a person is not exhausted by the quantity of output they can produce.
Part 3 - When Employment Becomes Obsolete
There is, however, a more difficult possibility hidden inside the problem. What happens if AI eventually becomes capable of performing so much economically useful work that hiring human beings for ordinary productive tasks is no longer necessary? It is possible to imagine a future in which the traditional economic justification for employment gradually weakens. If a machine can perform a task more cheaply, more quickly, and more reliably than a person, there may be little economic reason for a corporation to employ someone to do that task simply because humans have historically occupied that role.
This would create an unusual situation. A society could become extraordinarily productive while simultaneously reducing the amount of human labor required to sustain that productivity. The technological problem of producing goods and services could become easier while the social problem of distributing access to those goods and services became more difficult. The disappearance of necessary labor would not automatically eliminate people's need for resources, social participation, purpose, recognition, or a meaningful place within society.
One possible response would be some form of universal income or another mechanism for separating access to basic material security from participation in the labor market. Such a system would address one part of the problem: people would not necessarily need to sell their labor in order to survive. But income alone would not necessarily answer the deeper question of what people are expected to do in a society where their labor is increasingly unnecessary.
Employment performs several functions simultaneously. It produces things, but it also organizes time, creates social relationships, establishes status, develops skills, imposes responsibilities, and gives people a recognized role within a larger institution. If technological progress removes the productive necessity for human employment without replacing these other functions, the result could be a society that has solved a problem of material abundance while leaving a problem of human participation unresolved.
Part 4 - The Social Function of Work
This raises the possibility that future societies may deliberately preserve forms of human employment even when machines could perform the associated tasks more efficiently. Such employment would not necessarily exist because the human worker is the superior producer. It could exist because participation itself has social value. Corporations might be expected to contribute not only to production but also to the stability of the society from which they draw their customers, infrastructure, institutions, and workforce.
This would represent a significant change in the purpose of employment. A company might employ people partly because keeping people engaged in meaningful activity has social value, even when automation could accomplish the underlying work more efficiently. The economic value of the arrangement would therefore have to be understood more broadly than immediate output. A person might contribute through judgment, responsibility, creativity, institutional knowledge, mentorship, experimentation, or simply through participation in a structure that gives them a meaningful role.
This does not mean that corporations would necessarily have to manufacture meaningless jobs. Doing so could itself become a form of institutional dehumanization, merely replacing the statement "AI has made you unnecessary" with a different one: "We have invented a task so that you have something to do." The more interesting possibility is to find forms of participation that have genuine value even when AI performs the majority of routine production. The objective would not be to pretend that human beings remain economically indispensable. It would be to recognize that economic indispensability and social significance are not the same thing.
Part 5 - The Human Reserve
One particularly interesting possibility is the development of a human capability reserve. In such a structure, people would be organized partly according to specialized knowledge and exceptional abilities that could be maintained as a fallback against technological failure. An organization might normally rely heavily on AI for engineering, logistics, analysis, medicine, manufacturing, administration, or other forms of work, while maintaining a population of humans capable of understanding and performing those functions independently.
The rationale would not be that these people can routinely outperform the AI. Quite the opposite may be true. Their value would come from redundancy. Highly capable technological systems can create new forms of dependency, and dependency creates vulnerabilities. AI systems could become unavailable because of infrastructure failures, cyberattacks, corrupted data, technical malfunctions, supply-chain disruptions, deliberate interference, or other circumstances that cannot be fully anticipated. A society that has allowed every human capability to atrophy because machines can perform it better may discover that it has also eliminated its ability to recover when those machines stop working.
Human specialization could therefore acquire a new kind of value. Instead of being valued exclusively according to how much output a person can produce relative to an AI system, a person might also be valued according to the capability they preserve within the larger system. An engineer who rarely needs to intervene because an AI system handles most engineering tasks might nevertheless represent an important reserve of independent technical knowledge. A physician, pilot, scientist, technician, or administrator might similarly possess capabilities that are economically redundant under normal conditions but strategically valuable under abnormal ones.
The resulting hierarchy would consequently look different from a traditional corporate hierarchy. It might be based less on the number of people a manager can supervise or the amount of routine work an employee can complete, and more on specialization, judgment, experience, adaptability, and the difficulty of reconstructing a particular human capability after it has been lost. The organization would effectively maintain a living backup system.
This arrangement would also change the meaning of redundancy. In an ordinary business context, redundancy usually means unnecessary duplication. In a complex technological civilization, however, redundancy can be a form of insurance. Maintaining a capability that is rarely needed may appear inefficient until the moment that capability becomes indispensable. Human beings could therefore acquire value not despite their inability to compete with AI, but because maintaining an independent human capability provides a form of resilience that an AI-dependent society might otherwise lose.
Part 6 - A Society Beyond Economic Necessity
The deeper problem is that technological progress may eventually force society to distinguish between three things that have historically been closely connected: production, income, and participation. For most of industrial history, people could obtain income by producing something that another person wanted. Employment therefore served simultaneously as a mechanism of production, distribution, and social organization. Advanced automation could break that connection.
If machines eventually produce most of what society needs, there may be less reason to distribute resources primarily through wages. Yet eliminating the need for wages would not eliminate the need for a social structure in which people can develop abilities, exercise responsibility, form relationships, and participate in collective projects. The institutions of such a society would therefore have to be designed rather than inherited from an economic world in which human labor was indispensable.
This is ultimately why the question of comparing human beings with AI is more important than it initially appears. The danger is not simply that someone might be rude enough to say that a machine is smarter, faster, or more capable than a person. The deeper danger is that a society could begin treating those comparisons as the foundation of its institutions. If economic value becomes the only recognized form of value, then sufficiently capable machines could make an increasing number of human beings appear economically worthless.
A different approach would recognize that technological capability and human worth answer different questions. AI can determine what a machine is capable of doing. Economics can determine what activities are profitable. Neither question, by itself, determines what obligations people have toward one another or what kind of society they should construct. If technological progress eventually makes human labor largely unnecessary, the appropriate response may not be to force humans to compete indefinitely with machines, nor to pretend that machines have not changed the economics of work. It may be to build institutions in which material security, meaningful participation, and human capability can survive the disappearance of economic necessity.
Such a society might include universal income, voluntary and socially valuable forms of work, reduced working hours, institutions devoted to education and human development, and organizations that deliberately maintain human capabilities as a technological reserve. None of these possibilities is guaranteed to work, and each raises difficult questions about incentives, freedom, dignity, governance, and the distribution of power. But they share one important premise: the purpose of human beings in a technologically advanced society cannot be reduced to defeating the machines they created.
The paradox may ultimately be that the more successful AI becomes at doing what humans once had to do, the more important it becomes to decide what humans are for when they no longer have to do those things. The answer cannot simply be that humans must find some way to remain more productive than the machines. If that were the standard, technological progress would eventually turn every improvement in AI into another reduction in human worth. A more durable society would instead recognize that productivity is one measure of capability, not a complete measure of value.
References
- UNESCO (2021). Recommendation on the Ethics of Artificial Intelligence. UNESCO. 🔗
- International Labour Organization (2023). Generative AI and Jobs: A Global Analysis of Potential Effects on Job Quantity and Quality. International Labour Office. 🔗
- Acemoglu, D., & Restrepo, P. (2019). “Automation and New Tasks: How Technology Displaces and Reinstates Labor.” Journal of Economic Perspectives, 33(2), 3–30. 🔗
- NIST (2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0). National Institute of Standards and Technology. 🔗
- European Union (2024). Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act). Official Journal of the European Union. 🔗
