Where This Article Fits This article applies the Agency–Process–Scale (APS) framework to a long-term, explicitly speculative question about human evolution in artificial environments. It extends the APS treatment of biological agency, organised persistence, and evolutionary transformation without proposing new mechanisms of evolution or predicting particular inherited human traits. It distinguishes physiological maintenance, development, technological assistance, cultural transmission, and biological evolutionary change. Its purpose is to demonstrate how APS can organise questions about humanity’s possible future while preserving the explanatory roles of established evolutionary, physiological, and social sciences. The scenarios are conditional illustrations, not empirical forecasts or evidence that APS offers explanatory superiority over existing integrative frameworks.

Introduction: The Future of Human Evolution

How might artificial intelligence alter the conditions of human life, the organisation of human agency, and the evolutionary processes through which human populations change? What might human beings be like three thousand years from now?

The question invites familiar images of the future: larger brains, longer lives, genetically enhanced bodies, artificial organs, or an eventual merger of human intelligence with machines. Such possibilities are interesting, but they can also obscure a fundamental distinction. Changes in human biology, changes in human ways of living, and changes in the technologies supporting human life are not necessarily the same kind of evolutionary transformation.

Human biological evolution has not ended. Genetic variation, inheritance, reproductive differences, migration, and natural selection continue to influence human populations. But humans also modify the conditions under which these processes operate. Agriculture, medicine, sanitation, urbanisation, and modern communication have already transformed the environments in which human lives are maintained and reproduced.

Artificial intelligence may greatly extend this capacity.

Over the next three millennia, AI could become involved in nearly every major activity through which humans sustain their lives: obtaining food, maintaining health, acquiring knowledge, making decisions, raising children, organising communities, and managing environmental conditions. Some artificial systems may eventually operate with considerable independence, while others remain closely supervised or incorporated into larger social institutions.

The central question is therefore not whether AI will make human beings more intelligent or biologically superior. It is how increasingly artificial environments might alter the conditions under which human biological organisation persists, develops, reproduces, and evolves.

The Agency–Process–Scale (APS) framework provides one way of organising this question.

APS interprets biological agency as the viability-oriented activity of living organisation. Agency, Process, and Scale are complementary analytical perspectives on that organisation, not separate causal forces or successive levels of biological reality.

Applied to humanity’s possible future, APS encourages three connected questions:

  • Agency: How might AI alter the conditions under which humans maintain viable lives and exercise biological agency?
  • Process: Which physiological, developmental, behavioural, and technological processes might become involved in sustaining human organisation?
  • Scale: How might relationships among individuals, populations, institutions, technological systems, and evolutionary histories change?

These questions do not predict a particular future. They provide a framework for examining several possible futures while distinguishing biological change from the technological and cultural conditions that influence it.

1. Agency: Changing the Conditions of Human Life

Human agency does not begin with conscious decision-making. From an APS perspective, it is grounded in the organised biological activity through which living systems sustain the conditions of their persistence.

Human agency includes metabolism, physiological regulation, development, repair, behaviour, and interaction with the environment. Conscious planning and deliberate choice are important human capacities, but they are not the foundations of biological agency itself.

This distinction matters when considering artificial intelligence.

An AI system may advise someone about nutrition, detect a developing illness, manage an artificial organ, or coordinate the delivery of essential resources. Such activities may substantially affect the person’s ability to remain alive and function.

Yet the AI system’s contribution to human viability does not automatically make it part of the human organism, nor does it establish that the system possesses biological agency.

The human organism continues to maintain itself through materially realised biological processes. Artificial systems may alter the conditions under which those processes operate, supply external regulatory assistance, or become incorporated into particular physiological arrangements.

The relationship is therefore one of potentially increasing technological dependence and organisational coupling, not necessarily the creation of a single human–machine biological agent.

From Independent Regulation to Technologically Supported Regulation

Consider the historical development of medicine.

Human survival once depended much more directly on individual resistance to infection, environmental exposure, and nutritional scarcity. Vaccination, antibiotics, surgery, and public health systems have altered these conditions. They have not eliminated biological vulnerability, but they have changed its consequences.

AI-assisted medicine could extend this transformation through earlier diagnosis, personalised treatment, improved prostheses, automated monitoring, and more precise interventions.

A person with a serious physiological impairment might remain viable through continuous technological assistance. A population might sustain forms of biological variation that would have had different consequences under earlier environmental conditions.

In APS terms, the biological organisation remains the relevant bearer of human viability, while its conditions of persistence increasingly depend on externally organised technological processes.

This distinction also reveals a new vulnerability.

If continued functioning depends on artificial infrastructures, then failures of those infrastructures may become biologically significant. Energy supply, communication networks, medical technologies, and institutional reliability may become essential conditions of life for many individuals.

Greater technological support may therefore increase some forms of biological security while creating new dependencies.

Human agency could become more extensively supported without becoming more autonomous in every relevant sense.

2. Artificial Intelligence and the Changing Conditions of Natural Selection

One tempting prediction is that AI will reduce natural selection by protecting humans from environmental hazards.

This is too simple.

Natural selection occurs when heritable variation contributes to differences in reproductive success under particular conditions. Reducing mortality from one cause does not eliminate selection, and survival alone is not equivalent to evolutionary fitness.

AI-mediated medicine might reduce some differences in survival while changing reproductive patterns, fertility decisions, developmental conditions, and population structures. The evolutionary consequences would depend on how these changes affect inheritance and reproduction across generations.

Some previously disadvantageous inherited traits might become less consequential under improved medical care. Other traits could acquire different consequences in technologically transformed environments.

However, no particular genetic direction follows automatically.

Reproduction and Genomic Intervention

Reproductive technologies could become an especially important influence on future human populations.

AI might improve the interpretation of genetic information, support reproductive medicine, or assist the identification of some inherited disorders. More speculative possibilities include increasingly sophisticated embryo selection or heritable genome editing.

These possibilities require careful distinctions.

Preventing a particular inherited disease is not equivalent to producing a generally healthier or more intelligent population. Many human characteristics arise through complex interactions among numerous genetic variants, development, environment, and social conditions.

Traits such as intelligence, emotional regulation, and behavioural flexibility cannot responsibly be treated as simple genetic targets whose long-term consequences are predictable.

Nor should disability be equated with an absence of agency. A person’s capacities, biological vulnerabilities, and opportunities for living depend partly on social and technological conditions, not simply on an individual genetic characteristic.

Over three thousand years, intentional genetic interventions could influence the distribution of some inherited traits. Whether they become widespread would depend on technical feasibility, safety, reproductive choices, ethics, regulation, and unequal access.

If such interventions produce heritable changes that persist through generations, they would contribute to biological evolutionary transformation. But technological alteration of an individual’s body would not, by itself, constitute inherited evolutionary change.

APS helps preserve this distinction by connecting organised persistence with inheritance, variation, and differential reproductive outcomes without treating them as interchangeable.

3. Process: Reorganising the Maintenance of Human Life

Human life is maintained through coordinated processes extending across different spatial and temporal contexts.

These include cellular metabolism, immune defence, development, physiological regulation, behaviour, and social interaction.

Artificial intelligence could modify the environments in which these processes occur and the technological systems through which they are supported.

Physiological Regulation

Future healthcare might involve continuous monitoring of cardiovascular, metabolic, endocrine, and neurological conditions.

Some interventions could operate through external devices; others might involve implanted technologies, engineered tissues, or increasingly sophisticated artificial organs.

The resulting changes could be profound.

A person might survive with physiological conditions that would previously have been incompatible with prolonged life. Chronic diseases might become more manageable. Some processes of ageing could become more amenable to intervention.

But these developments would not necessarily mean that natural selection had produced humans biologically adapted for longer life.

They might instead represent technological extensions of the conditions supporting existing biological organisation.

This is an important APS distinction: an improvement in the maintenance of an individual’s viable organisation is not automatically an evolutionary adaptation.

The same distinction applies to longevity.

If future humans routinely live longer, the explanation might involve medicine, environmental protection, altered behaviour, inherited biological changes, or combinations of these factors. Different explanations would require different evidence.

Development and Plasticity

Human development is responsive to environmental conditions.

Nutrition, social interaction, education, stress, disease exposure, and cultural practices influence how inherited capacities are expressed.

AI-mediated environments could substantially change these developmental conditions.

Children might learn through highly personalised systems. Language acquisition, technical competence, and access to knowledge could be supported by artificial tutors. Cognitive tasks that once required extensive individual training might be performed through continuous interaction with external systems.

Such developments could change human abilities without requiring corresponding genetic evolution.

A person might become capable of solving problems that would previously have exceeded their unaided capacity, while the underlying biological organisation remained broadly similar.

This would represent a transformation in the organisation of human activity rather than necessarily a transformation in inherited human cognition.

APS encourages us to ask how the relevant processes are materially realised and maintained, rather than assuming that every enhancement of performance represents a new biological adaptation.

4. Cognition in an Artificially Structured Environment

Human cognition is already deeply dependent on external resources.

Language, writing, mathematics, maps, scientific instruments, and computers extend the possibilities of memory, reasoning, and communication.

AI may greatly increase the sophistication of these resources.

Future humans could routinely use artificial systems to organise information, explore alternatives, translate between languages, identify patterns, and support decisions.

The resulting cognitive environment might be unlike anything familiar today.

Nevertheless, three different possibilities must be distinguished.

First, human biological cognition might remain substantially similar while artificial systems provide increasingly powerful external assistance.

Second, development within AI-rich environments might alter acquired cognitive skills, habits of attention, and forms of social learning.

Third, inherited human cognitive characteristics might change across generations if relevant genetic variation produces reproductive differences under these conditions.

These possibilities are compatible, but none entails the others.

It would therefore be premature to predict genetically enhanced working memory, greater attentional flexibility, or increased emotional stability as inevitable outcomes.

Indeed, extensive technological assistance could reduce the practical importance of some unaided capacities while increasing the importance of others.

The consequences might also differ across populations and social environments.

Biological Cognition and Artificial Computation

APS distinguishes biological agency from cognition and does not treat every adaptive response as cognitive.

Human cognition involves biological organisation, but artificial computation can contribute to human cognitive activity without automatically becoming biological cognition.

Whether a particular artificial system warrants attribution of agency or cognition is a separate question requiring explicit criteria.

This distinction is especially important when discussing future human–AI integration.

An implanted computational device might participate in a person’s behaviour or neural regulation. But determining whether it belongs to the organisation of the biological agent, functions as an external aid, or constitutes a distinct interacting system would require analysis of its actual dependencies.

Technological proximity alone does not settle biological individuality.

5. Scale: Human Life Across Biological, Social, and Technological Contexts

The third APS perspective concerns the extent of the dependencies relevant to biological organisation.

Human life is not sustained through isolated physiological activity.

Food production, water supplies, sanitation, healthcare, education, family relationships, and political institutions all influence the conditions under which human organisms persist.

AI could alter these dependencies on a scale extending far beyond individual bodies.

Consider food security.

An individual’s metabolism depends on nutrients, but access to those nutrients may depend on agricultural technologies, transport networks, economic institutions, environmental conditions, and increasingly AI-mediated coordination.

The biological process remains metabolism. The conditions supporting it extend through a much wider organisation of activities and resources.

APS can describe these relationships without claiming that a food-distribution network is itself a biological organism or that every participating institution possesses biological agency.

The distinction between the biological bearer and its supporting conditions remains essential.

Social Organisation and Cooperation

AI could transform how communities coordinate labour, distribute resources, manage conflict, and make collective decisions.

These changes might influence human behaviour and cultural evolution.

However, it does not follow that genetic selection will necessarily favour more cooperative, compliant, or emotionally stable individuals.

Social behaviour is shaped by complex interactions among development, culture, institutions, individual differences, and inherited dispositions.

The reproductive consequences of particular behavioural traits would depend on historically contingent circumstances.

AI-supported societies might reward cooperation in some settings and competition in others. Different institutions might encourage very different patterns of behaviour.

Moreover, what benefits an institution is not necessarily what benefits an individual organism or increases evolutionary fitness.

APS therefore requires us to specify whose viability, persistence, or reproductive success is under consideration.

A technologically efficient society is not automatically a biologically beneficial one.

6. Cultural Evolution and Technological Inheritance

Across three thousand years, changes in cultural and technological inheritance may be much faster and more extensive than changes in human genetic composition.

Human societies transmit knowledge, institutions, skills, technologies, and practices between generations.

AI could accelerate these processes by preserving information, generating new technical possibilities, and assisting the reproduction of complex systems.

A future child might inherit not only genes and a social environment, but access to highly developed artificial infrastructures created over many generations.

Such inheritance can have profound biological consequences.

Agriculture altered human diets and disease environments. Urbanisation transformed patterns of exposure, reproduction, and social interaction. Future artificial environments might similarly change the circumstances in which human biological variation matters.

But cultural transmission and genetic inheritance are not identical processes.

Technological systems can be copied, redesigned, or abandoned without corresponding genetic change. Cultural practices can spread horizontally between unrelated individuals. Their persistence may depend on institutions, communication, and material infrastructures rather than biological reproduction alone.

APS can help organise the dependencies among these forms of continuity, but it does not replace the established evolutionary explanations appropriate to each.

A convincing account of future human evolution must therefore distinguish biological inheritance, cultural transmission, technological reproduction, and their interactions.

7. Four Possible Human Futures

Rather than predicting a single evolutionary destination, we can consider four contrasting possibilities.

Scenario A: Biological Continuity with Extensive Technological Support

Human genetic and physiological characteristics remain broadly recognisable, while medicine, artificial intelligence, and engineered environments transform everyday life.

Many improvements in health and performance arise through external assistance rather than inherited biological change.

The principal transformation concerns the conditions under which human agency operates.

Scenario B: Uneven Biological and Technological Modification

Some populations gain access to advanced reproductive interventions, implants, engineered tissues, and specialised medical systems, while others do not.

Human biological variation becomes influenced by unequal technological access, reproductive practices, and social institutions.

The consequences could include new patterns of vulnerability, dependence, and inequality.

However, social differentiation would not necessarily amount to the formation of separate biological species.

Scenario C: Deep Physiological Integration

Artificial devices become increasingly integrated with physiological regulation.

Some people depend on implanted systems for sensory processing, neural function, circulation, or metabolic control.

The distinction between organism and supporting technology becomes more difficult to analyse in particular cases.

APS would require careful investigation of the material and organisational relationships involved rather than assuming that integration automatically creates a new biological individual.

Scenario D: Technological Disruption and Renewed Ecological Constraints

Advanced artificial systems fail, become inaccessible, or are disrupted by environmental, political, or infrastructural crises.

Human populations face renewed pressures associated with resource scarcity, disease, environmental change, and institutional instability.

Under these circumstances, previously reduced vulnerabilities could regain importance.

This scenario reminds us that technological dependence does not guarantee permanent security.

These four possibilities are not mutually exclusive. Different populations could experience different combinations over the same historical period.

8. What Would Count as Human Evolution?

The phrase human evolution is often used loosely to describe any major transformation in human life.

APS encourages greater precision.

A future change in human behaviour might be developmental or cultural. A change in survival might result from medical intervention. A change in population characteristics might reflect migration, demographic processes, natural selection, or intentional reproductive choices.

Biological evolution requires changes in inherited characteristics of populations across generations.

The relationship between these changes and the persistence of living organisation must be established rather than assumed.

This matters particularly when imagining a future human–AI hybrid.

A human using artificial intelligence is not necessarily a biological hybrid in any strict sense. Nor does dependence on technological systems create a new species.

Species formation would require much stronger evidence concerning evolutionary divergence and reproductive relationships.

Over three thousand years—approximately a hundred human generations if the average generation interval were thirty years—substantial biological change is possible under some circumstances.

But the timescale alone does not determine the direction or magnitude of that change.

The more defensible prediction is that human biological evolution will continue within environments increasingly shaped by technological and cultural processes, provided those infrastructures persist.

The particular consequences remain uncertain.

9. What Does APS Contribute to This Question?

APS does not provide a mathematical model capable of predicting human traits in the year 5026.

Nor does it introduce a new mechanism of inheritance, selection, or technological change.

Its possible contribution is conceptual: organising questions that are easily confused when discussions of the future move between biology, technology, cognition, and society.

The Agency perspective asks which activities sustain human viable organisation and how artificial systems modify their conditions.

The Process perspective asks how those activities are maintained, transformed, disrupted, or restored through time.

The Scale perspective asks how the relevant dependencies extend across cells, organisms, populations, technological infrastructures, and historical processes.

Together, these perspectives encourage an account of human evolution that preserves the differences among biological persistence, developmental change, technological assistance, cultural inheritance, and evolutionary transformation.

Much of this explanatory work is also addressed by established approaches, including evolutionary theory, niche construction, developmental systems research, organisational autonomy, and cultural evolution.

APS should therefore not be credited with a unique explanatory contribution merely because it brings these subjects into a common discussion.

Its distinctive usefulness, if any, would have to be established through comparison with equally comprehensive alternative frameworks.

The present interpretation is best understood as an application of APS to a speculative future, not a demonstration of its explanatory superiority.

Conclusion: Humanity in the Year 5026

Humanity three thousand years from now may be biologically familiar yet extraordinarily different in the conditions of its existence.

Artificial systems might support health, reproduction, education, environmental regulation, and collective organisation to an extent that is difficult to imagine today.

Some human characteristics may change through genetic evolution. Others may change through development, cultural transmission, or direct technological intervention. Many apparent enhancements could arise from artificial systems without corresponding inherited biological modification.

It is possible that the most significant transformation will not be a new human anatomy or a uniformly enhanced intelligence, but a far-reaching reorganisation of the conditions under which human life is maintained.

Such a future would bring new possibilities and new vulnerabilities.

Technological assistance could expand the range of conditions compatible with viable human lives. At the same time, dependence on artificial infrastructures could create unfamiliar forms of biological risk.

From an APS perspective, the essential continuity would remain the activity of living organisation sustaining its conditions of persistence. What changes is the organisation of the processes and environmental relationships through which that persistence becomes possible.

Humanity in 5026 need not be a new species, nor must it become a single biological–technological organism.

It may instead consist of human populations whose biological lives are sustained within increasingly complex, historically accumulated systems of artificial support.

The central evolutionary question is not whether humans will become better adapted to machines. It is how inherited biological variation, organised persistence, cultural practices, and technological environments will continue to influence one another.

The future of human evolution may be shaped less by a predetermined direction of biological improvement than by the changing conditions under which human agency remains possible.

Suggested Reading

  • Dawkins, R. (1982). The Extended Phenotype. Oxford University Press.
  • Odling-Smee, F. J., Laland, K. N., and Feldman, M. W. (2003). Niche Construction: The Neglected Process in Evolution. Princeton University Press.
  • Richerson, P. J., and Boyd, R. (2005). Not by Genes Alone: How Culture Transformed Human Evolution. University of Chicago Press.
  • Moreno, A., and Mossio, M. (2015). Biological Autonomy: A Philosophical and Theoretical Enquiry. Springer. https://doi.org/10.1007/978-94-017-9837-2
  • Paul, A. M. (2021). The Extended Mind: The Power of Thinking Outside the Brain. Houghton Mifflin Harcourt.
  • Spencer, R. D. (2026). Agency as the defining activity of life: A viability-oriented framework integrating process and scale. Biological Theory. https://doi.org/10.1007/s13752-026-00547-6