The Thesis
Every civilization that has survived existential threat did so through one mechanism: learning. Not technology, which is inert without understanding. Not resources, which are useless without the knowledge to deploy them. Not institutions, which collapse when the people within them cannot adapt. Learning is the process through which human societies recognize change, generate responses, and transmit solutions across generations. It is, in the strictest sense, humanity’s only scalable adaptation technology.
I see this not as metaphor, but as mechanism.
Historical Context: The Learning Society Concept
The notion of a “learning society” is not new. Since the mid-1980s, policy makers, economists, and scholars have argued that investment in education and training constitutes the primary route to organizational and national advantage (Morris, 1999). This approach drew inspiration from economists dating back to Adam Smith and Alfred Marshall, as well as the human capital theory developed by Gary Becker in the 1960s. More recently, these ideas have been rebranded under various labels, including the “learning society” and “knowledge economy.”
Yet as Morris (1999) observed in his comprehensive review of the field, there remained “far too little comparable data available for robust conclusions to be drawn about the incidence and effects” of education and training investments across different national contexts. He called for “the development of more comprehensive and comparable surveys” to move the debate forward.
A quarter century later, this call remains largely unanswered, even as the stakes have multiplied exponentially with the emergence of artificial intelligence.
What was once a question of national economic competitiveness has become a question of civilizational adaptation. The learning society thesis, articulated in an era of globalization and human capital theory, now confronts a technological discontinuity that its original proponents could not have imagined: systems that learn without humans, that generate knowledge without understanding, that threaten to substitute for the very educational processes they were meant to enhance.
The Biological Lens: From Genes to Culture
Biological evolution adapts species to environments through genetic variation and natural selection. The process is powerful but slow, operating across generations, blind to intention, incapable of foresight. For most of life’s history, this was sufficient. Environments changed gradually; species either adapted or perished.
Humans broke this constraint. Somewhere in our evolutionary history, we developed a second inheritance system: culture. Unlike genes, cultural information can be acquired within a lifetime, modified deliberately, and transmitted horizontally across populations. A human born today inherits not merely the genes of ancestors but the accumulated knowledge of thousands of generations: language, mathematics, medicine, governance, the hard-won lessons of countless trials and catastrophes.
This is the first great insight: culture is a faster adaptation system than biology. When environments change rapidly, as they did during ice ages, as they do during pandemics, as they will during climate collapse, species constrained to genetic adaptation cannot respond in time. Humans respond through culture. We learn, we teach, we transmit, and we adapt at the speed of communication rather than the speed of reproduction.
But culture does not transmit itself. It requires infrastructure.
The Cultural Lens: Why Learning Requires Institutions
Cultural transmission happens through demonstration, through story, through apprenticeship, but none of these mechanisms scale. A master craftsperson can train a handful of apprentices. A storyteller can reach a village. Even writing, that great leap, requires readers who have been taught to read.
Education is the institution humanity invented to solve this scaling problem. Schools, curricula, pedagogies: these are not merely social conventions. They are the infrastructure through which culture becomes reliably transmissible across populations and generations. Without deliberate educational systems, knowledge fragments, erodes, disappears. With them, knowledge compounds.
Consider what happens when educational infrastructure fails. The Bronze Age Collapse erased writing systems across the Mediterranean for centuries. The fall of Rome fragmented technical knowledge that took a millennium to recover. The deliberate destruction of indigenous educational systems during colonization severed cultural continuity in ways that communities still struggle to repair. These are not merely historical tragedies. They are data points. They demonstrate that societies which lose the capacity to educate lose the capacity to adapt.
And here is the compounding insight: education does not merely transmit existing knowledge. It generates new knowledge. The capacity to learn how to learn, to develop methods for inquiry, criticism, and revision, is what separates societies that stagnate from societies that progress. David Deutsch captures this precisely: all progress depends on the creation of better explanations through conjecture and criticism. Education is the institution that cultivates this capacity at scale.
The Society & AI Lens: Amplifier or Threat?
Artificial intelligence represents a discontinuity in human history. For the first time, we have created systems that can process information, generate outputs, and increasingly, act in the world without direct human intervention. The question is not whether AI will transform society; it already is. The question is what that transformation means for humanity’s adaptation mechanism.
Two futures are possible. The design choices we make today will determine which one we inhabit.
In the first, AI amplifies education. It personalizes instruction, extends access, frees teachers from administrative burden, and accelerates the creation and dissemination of knowledge. In this future, AI is a tool that serves the learning society, making education more effective, more equitable, more capable of preparing humans to navigate whatever challenges emerge.
In the second future, AI substitutes for education. It generates answers without cultivating understanding. It provides outputs without developing the capacity to evaluate them. It creates dependence rather than capability. In this future, humans become consumers of AI-generated solutions rather than producers of knowledge. And when those AI systems fail, as all systems eventually fail, humans will have lost the very capacity that allowed our species to survive ice ages and plagues and civilizational collapse.
This is not a matter of technology policy. It is a matter of species survival.
The learning society thesis clarifies the stakes. If education is humanity’s adaptation mechanism, and if AI is being integrated into educational systems worldwide, then the design choices we make today will determine whether future generations can adapt to challenges we cannot yet imagine. We are not simply deciding how to teach mathematics or reading or science. We are deciding whether the infrastructure of human adaptation will be strengthened or eroded.
The Design Imperative
The learning society thesis is not merely analytical. It is prescriptive. It demands that we evaluate every AI intervention in education against a single criterion:
Does this cultivate human capacity for learning, or does it substitute for it?
AI that helps learners struggle productively with difficult problems cultivates capacity. AI that provides answers without struggle substitutes for it. AI that extends access to quality instruction for the billions currently excluded cultivates capacity at civilizational scale. AI that concentrates educational advantage among the already privileged substitutes global adaptation capacity for local efficiency gains. AI that supports teachers in understanding their students cultivates the relational core of education. AI that replaces teachers with interfaces substitutes technological optimization for human development.
The distinction is not always obvious. A system can appear helpful while subtly undermining the learning it claims to support. A system can feel demanding while genuinely building capability. The learning society thesis provides the evaluative lens: we must always ask whether humans are becoming more capable of adapting, or less.
We are a learning society; the question is whether we will remain one.
