The Chicken, the Egg, and the Algorithm

Here is a question humanity keeps stumbling over, regardless of the domain: Do we change people first, or the systems they inhabit? Do we fix the individual before fixing the institution, or the institution before the individual? The question sounds reasonable. De facto, it is a trap.

The chicken-and-egg conundrum is seductive because it offers an alibi. It lets individuals say, “My behaviour doesn’t matter until the system changes,” and lets institutions say, “We’ll act when people are ready.” In practice, both groups wait indefinitely, and the window of possibility quietly closes. This dynamic appears, with painful reliability, in two of the defining challenges of our time: the climate emergency and the accelerating integration of artificial intelligence (AI) into every layer of daily life.

Same Argument, Different Epochs

For decades, the climate debate has been cleaved into two camps that rarely speak honestly to each other. One says the problem is fundamentally structural: fossil fuel subsidies, extractivist economics, regulatory capture by carbon-intensive industries. The other says culture and individual behaviour are the levers—eat less meat, fly less, consume more deliberately and waste less. Both are right. Yet each tends to use its own correctness as a reason to dismiss the other.

A new article in Nature Climate Change cuts through this false binary. Climate discourse, the authors argue, too often frames individual behaviour and systems change as separate, competing pathways. Their alternative proposition: Social change arises from individuals exercising agency within societal systems, and that agency should be actively leveraged—not opposed to systemic work but woven through it. A parallel 2026 study measuring “carbon capability” reinforces the point, showing that when researchers measure only personal carbon footprints, they miss all the public-sphere contributions—advocacy, collective action, civic engagement—that make individual behaviour matter at scale.

The implication is not subtle. Investing in one side of the equation while neglecting the other is not a half-solution. It is a non-solution dressed up as pragmatism.

Same Trap, New Coordinates

The chicken-and-egg debate around climate change is more than three decades old. The AI version is just beginning, yet it is replicating and supercharging the same structural error.

Some argue that the real dangers related to AI are systemic: opaque algorithms, algorithmic bias, the concentration of AI capability in the hands of a few corporations, surveillance architectures that erode collective privacy, environmental burdens. Others insist the problem is individual: people who outsource their thinking, who accept recommendations without interrogation, who lose the muscle of independent judgment because they stopped exercising it, who establish relationships with chatbots at the expense of interpersonal connections.

Both camps are describing something acutely real. Research on agency decay documents how algorithmic dependence progresses through four stages—from curious exploration to integration, to reliance, to a form of learned helplessness where autonomous decision-making becomes genuinely difficult. This has begun to reshape organisations, creating feedback loops that erode both confidence and capacity. Meanwhile, systemic critiques point to how algorithmic architectures reshape identity, introspection, and moral agency—not through any single choice by any single person, but through cumulative environmental design.

The individual without the enabling system is a seed without soil. The system without the individual is fertile ground without farmers. Neither functions without the other.

A Framework Built on Simultaneity

The ProSocial AI Initiative was designed from its inception around this insight: that meaningful change in the human-AI relationship requires concurrent investment at the level of the person and at the level of the system. These are not parallel tracks that might one day converge. They are two hands of the same body.

• Change from the inside out: The individual side of this equation is addressed through double literacy—a framework that pairs two distinct but inseparable capacities. Human literacy means understanding and actively cultivating natural human intelligence in all its dimensions: emotional, relational, embodied, aspirational. Algorithmic literacy means understanding and critically engaging with the AI and digital systems that increasingly mediate our choices, relationships, and opportunities. Neither literacy is sufficient alone. Someone fluent in algorithms but estranged from their own inner life is as poorly equipped for the hybrid future as someone rich in self-knowledge but oblivious to how AI systems are shaping their environment and their own behavior. Education within the hybrid era must be reconceived as a collective civic practice, rather than a set of individual technical skills.
• Change from the outside in: A tool to address the systemic side of the equation is the ProSocial AI Index—a measurement framework designed to map, measure, monitor, and manage the algorithmic architecture of organisations, cities, and nations with regenerative intent. The index is built on the logic that what gets measured gets managed, and that current metrics—productivity, engagement, growth—systematically ignore the human and planetary costs of how AI is deployed. The ProSocial AI Index asks different questions: Does this system build or erode human capacity? Does it serve collective well-being or pure profit optimisation? Does it account for environmental impact and intergenerational equity?

These are governance questions, and they require governance-level responses. Individual awareness, however widespread, cannot redesign an algorithm. Algorithmic redesign, however sophisticated, cannot cultivate the judgment needed to use it wisely.

SPARK: A Practical Takeaway

Complexity does not preclude action. It reframes it. For anyone navigating the hybrid future—as a parent, educator, manager, policymaker, or simply a person trying to think clearly in an AI-saturated world—the following five moves offer a starting point:

Artificial Intelligence Essential Reads

AI and the Psychology of Cognitive Surrender

AI Officially Passes the Turing Test, Landmark Study Shows

S — Start with yourself. Before evaluating any AI tool or system, invest five minutes in noticing your own state: your assumptions, your mood, your biases. Natural intelligence is the instrument through which you interpret everything else. Attend to it deliberately.

P — Probe the system. Ask, of any AI system you use regularly: What does it optimise for? Whose values shaped it? Who benefits from its design? You do not need to be a data scientist to ask these questions. You need to be a citizen.

A — Act at both levels simultaneously. Refuse the either-or. Change your own habits and advocate for structural reform. Support your own child's critical thinking and push for algorithmic transparency in the schools and platforms they inhabit.

R — Resist the learned helplessness narrative. Agency decay is real, but it is (still) reversible. Skills that atrophy can be rebuilt. The first step is refusing the story that individual action is futile in the face of large systems—a story, incidentally, that large systems find very convenient.

K — Keep the long view. Neither a person nor an institution transforms overnight. The hybrid future will be shaped by thousands of small, simultaneous decisions—about what to cultivate in ourselves, what to demand from our technologies, and what kind of intelligence we want to inhabit the world we are building. Those decisions are being made right now. The question is only whether they are made deliberately.

The chicken and the egg were always a false problem. Both exist because both were always nurtured and necessary. The same is true of individual and systemic change—and of the two literacies required to navigate a world where natural and artificial intelligences coexist.