The Age of Agency: A World Where Building Is No Longer Optional
Every generation inherits tools. Only a few inherit agency.
The Line Has Been Drawn
It’s worth framing the core reality of the moment we are in:
The systems that once buffered failure, such as standardized curricula, age-based cohorts, and credential gates, no longer function as safety nets. When learning becomes infinite and on-demand, progress depends almost entirely on agency. I believe that this shift explains more about the present, and the coming divide, than almost any headline about AI.
I’ve been here before.
In 2015, I wrote about a world where anyone, anywhere could shape the world around them. It was a hopeful piece, grounded in stories of young people, often far from traditional power centers, discovering that with the right tools, curiosity could become creation, and creation could become confidence. The internet was beginning to flatten opportunity and the ladder still existed, albeit widening.
At the time, it felt inevitable that access would lead to mobility. That if you could learn, you could rise. That connectivity would translate, almost automatically, into progress. A decade later, that ladder is gone. Nearly 40% of recent U.S. college graduates are underemployed, working in jobs that do not require a degree, a signal that credentials no longer guarantee trajectory [1].
As I argued in The Vanishing Ladder, the familiar sequence of education to career, career to stability, stability to meaning, no longer functions the way we pretend it does. Worth stating that the system didn’t collapse in a crisis and didn’t fail all at once. Instead, they’ve been dissolving quietly while we kept handing out diplomas, offering internships, and reassuring people that the path was still there for anyone willing to work hard enough.
I don’t believe that AI is causing this collapse but it will certainly rapidly remove the illusion that it could be repaired. What AI is doing now is more unsettling. It is drawing a line between those who shape systems, and those who are shaped by them. Between builders and consumers. You can think of it as a sorting function, one already underway (and accelerating).
Learning Has Changed Shape
For most of the industrial age, learning happened inside institutions. Schools decided what mattered, degrees signaled readiness, and progress moved step by step, credential by credential. This structure made sense when knowledge was scarce, expensive to distribute, and slow to change.
It makes little sense now. Today, intelligence is ambient. Think AI tutors, copilots, simulators, and conversational systems that will soon sit in the pockets of millions of students. UNESCO estimated that by the middle of this decade, a majority of learners in advanced economies will use some form of AI-assisted education [2].
This is often framed as a story about access, but access is no longer the story. The real shift is quieter, and more uncomfortable, and that is access is no longer the differentiator, but intent is. Two people can sit in front of the same tools. One will use them to explore, to test ideas, to interrogate assumptions, to build mental models of how the world works. The other will use them to skim, comply, and stay entertained. One compounds capability while the other depresses curiosity.
AI won’t reward effort in the way schools once did. It won’t care how long you studied or how closely you followed the syllabus. Instead, it will respond to the quality of questions, the sharpness of feedback loops, and the willingness to revise.
Learning is not disappearing. It is simply becoming directional. Meaning, there will no longer be a shared path, only self-chosen vectors. There won’t be default curriculum that carries people forward, and no cohort that ensure people progress simply by showing up. Learning will amplify whatever direction people will point themselves in. If they aim it toward curiosity, experimentation, and synthesis, it will compound rapidly. If they aim it toward entertainment, validation, or passive consumption, it will do the same.
I believe that this shift will expose a deeper truth about education that predates AI, that institutions were never designed primarily to maximize learning, but were actually designed to sort people into roles, hierarchies, and systems of legitimacy. Meaning, credentials were signals, not guarantees of capability.
AI is breaking this model by making knowledge abundant and verification continuous. We will no longer need permission to learn, but we will also no longer have protection from falling behind. The systems that once buffered failure, such as standardized curricula, age-based cohorts, credential gates will no longer function as safety nets. Instead, learning will be infinite and on-demand, which means that progress will depend almost entirely on agency.
It’s really quite a profound shift, one we are deeply unprepared for it.
Making Is Becoming the Baseline
The maker movement fostered the approach that people learn best by building, anchored on the work by Seymour Papert, pioneer of constructionist learning. AI won’t replace that insight. Instead, it completes it.
In The Age of Replacement, I wrote that AI won’t just automate tasks. It will replace the logic of work itself, because when execution becomes cheap, fast, and repeatable, the economic value of following instructions will disappear. What could remains valuable is deciding what should exist in the first place.
We are watching this happen in real time. Microsoft has publicly stated that roughly 30% of its internal codebase is now written with AI assistance, not as an experiment, but as production infrastructure [3]. This means that execution, once a durable moat, is rapidly becoming a commodity, and the implication is not that humans are becoming obsolete, but that a certain kind of human is.
For decades, we trained people to excel at executing predefined tasks. We rewarded compliance, specialization, and efficiency. That made sense in a world where humans were the primary engines of execution. AI is flipping that equation. When machines can execute, value is shifting toward framing problems, defining constraints, exercising judgment and taste, and deciding not just how something works, but whether it should exist at all.
This raises the question of how would people acquire judgment, taste, and responsibility in a world where traditional education is fragmenting? I believe that the answer is not more information, but guided creation.
Judgment emerges from seeing consequences. Taste develops through iteration and exposure to quality. Responsibility forms when creation affects others, not just grades or metrics. These qualities cannot be downloaded. They must be practiced, and that’s why making will no longer be a niche skill reserved for engineers or designers. It will become a form of literacy, which is why the new divide won’t be between technical and non-technical people, but between those who can translate intent into reality, and those who cannot.
Builders don’t need to be perfect nor fast, but they do need to act. Builders test ideas against reality, learn from feedback, and refine understanding through creation. Those who cannot build, even at a conceptual level, will become dependent on systems they do not understand and cannot influence. Which means that dependency will become the default, and agency will become the exception.
This could become the baseline of the AI age.
Sharing Is About Contribution Again
The social internet (roughly the period between 2010-2025) trained us to share for attention. Likes, views, and followers. It was a visibility economy, optimized for reach rather than value. The next phase of the internet could be much quieter, and far more consequential.
Over 90% of the world’s digital infrastructure now relies on open-source software, maintained by distributed communities of builders [4]. These systems are governed by contribution, not credentials or titles. AI will accelerate this model as it will lower coordination cost, translate across languages, generate documentation, and enable global collaboration at unprecedented speed. Power will flow through participation, not institutions or resumes. Reputation will accrue to those who build things others rely on, and legitimacy will be earned through usefulness.
I believe that this shift will reach far beyond software. We are watching early versions of new institutions emerge, such as parallel systems for learning, creating, funding, and governing, built by networks of contributors rather than centralized authorities. In software, this shift is already visible with platforms like GitHub which function as parallel education systems, where learning happens through contribution rather than coursework, assessment happens through code review rather than exams, and credentials are earned publicly through sustained participation. Millions of developers now build careers by shipping code others rely on. In capital formation, platforms like AngelList have done something similar to venture funding but essentially unbundling it from elite firms and enabling capital to flow through individuals, track records, and networks. These example systems are operational, widely used, and already reallocating opportunity away from centralized authorities toward networks of contributors.
And so, consumers will scroll, builders will build and compound, and the gap between them will widen.
The Inequality We Avoid Naming
We spend a great deal of time talking about access to the internet, to tools, to opportunity, but access is table stakes. More than two-thirds of the world is now online [5], yet only a small fraction of internet users (on the order of a few percent) meaningfully contribute to open-source software, digital products, or the systems that shape the digital economy, as reflected in GitHub’s annual Octoverse report. [6]. This means that meaningful participation in the future economy remains concentrated among a relatively small group.
Why? Because access without agency produces consumption, not creation. Without the skills, confidence, and permission structures to act, abundance will become overwhelming rather than empowering, attention will fragment, and tools will become entertainment. The missing variable is agency.
Agency is the capacity to decide what to learn, what to build, and what responsibility to take for outcomes. I also see it as the the willingness to move from consumption to creation, from opinion to ownership. AI will amplify this difference. Which means that for those with agency, it will be a leverage but for those without it, it will become sedation. Again, one should see it as a sorting mechanism of the AI age.
Institutions Are Losing the Plot
The most destabilizing force of the AI age is institutional, not technological. For centuries, institutions existed to coordinate people at scale, legitimize progress, and reduce uncertainty. Schools translated curiosity into credentials, corporations translated labor into wages, and governments translated participation into belonging. AI could disable educational institutions faster than almost anything else, and not because education is unimportant, but because the problem education was built to solve has fundamentally changed.
For centuries, these institutions existed to manage scarcity: scarce knowledge, slow feedback, rigid sequencing, and the risk of falling behind. AI will collapse all of that, delivering explanation, practice, and feedback directly, continuously, and at marginal cost, while stripping away the sequencing, credentialing, and failure-buffering functions that once justified institutional mediation. With learning becomes infinite and on-demand, institutions will no longer control progress, and will no longer protect individuals from disengagement. What’s likely going to break is not the value of education itself, but the delivery model that assumed learning needed to be centralized, paced, and permissioned.
This raises two questions we now have to confront: What is the value of education? And what is the value of creation?
I believe that education’s true value is not information transfer but orientation, helping people decide what matters, how to think, and how to learn continuously. On the other hand, creation’s value is not necessarily output, but agency, which entails learning through impact. Institutions that fail to deliver these will lose legitimacy.
Alternative education paths, creator-led companies, network-native capital, and open-source governance are essentially early versions of replacement institutions, ones that are lighter, faster, and more accountable to contribution than to credentials. This is what institutional decline looks like in real time. They don’t even need to collapse, they are simply becoming irrelevant.
Agency Is a Moral Choice
It’s tempting to frame the builder–consumer divide as a matter of talent or privilege, but in reality, it’s not. At its core, agency is a moral choice. When creation is easy, some things become dramatically easier to build, such as software, media, and prototypes, experiments, while other things become harder, such as wisdom, trust, meaning, and responsibility. Choosing what to build, and what not to build, will become a defining act.
When learning becomes infinite, choosing not to learn will become a posture, and when contribution is optional, opting out will become an identity. This is uncomfortable because it will remove excuses, and there will no longer be a single gatekeeper to blame, no syllabus to hide behind, and no institution to wait for permission. The tools will be present and the leverage will be available. What will essentially remains is choice.
Builders always, and will continue to choose discomfort and uncertainty. They will test ideas against reality rather than opinions against timelines. Consumers on the other hand will choose ease and familiarity. They will let systems decide what is worth attention, effort, and care. Each choice will compound very differently over time. The future won’t punish consumption but will simply reward building useful things, meaningful systems, and shared value.
A Forked Childhood
Children growing up today will not remember a world without AI. They will not experience it as disruption but instead, as an environment. Early studies suggest that students using AI-assisted learning environments can reach mastery of complex problem-solving tasks 20–30% faster when guided well, typically by structured pedagogical frameworks, mentors, or adaptive feedback systems rather than by the models alone [7].
And so guidance matters more than the technology itself. Left on its own, AI will drift toward convenience and consumption, answering, entertaining, and smoothing over effort. But when guided with intent, it could becomes something else entirely: a tool for exploration, creation, and judgment. Important to note that this will shift responsibility to parents, educators, and leaders in ensuring that we’re not shielding the next generation from technology, but instead, modeling agency in a world where opting out of responsibility is easier than ever.
Why This Is Personal
I think about this future through my daughter. She is growing up in a world where intelligence is becoming ambient and creation is becoming conversational. AI will feel to her the way electricity felt when it became widespread: invisible, assumed, and can be found everywhere. I believe that what will shape her and her generation’s future is not whether she will have access to powerful tools but whether she will believe she has the right, and the responsibility, to build things that move the human condition forward.
To be clear: not everyone needs to build. Societies will still need caretakers, stewards, and consumers, but everyone will need agency, the ability to participate meaningfully in shaping the systems we depend on.
Kano was born from that belief. I watched children around the world change the moment they realized they could shape technology instead of just consuming it. Confidence followed creation and identity followed agency. That moment is now arriving at civilizational scale.
Despite what many claims, AI does not guarantee a better future but it certainly forces us to make a clearer choice between becoming a builder or remaining a passive consumer. That is the line, and this time, it is unmistakable.
With belief,
Yon
References
[1] Federal Reserve Bank of New York, Labor Market for Recent College Graduates. https://www.newyorkfed.org/research/college-labor-market
[2] UNESCO, Global Education Monitoring Report: Technology in Education, 2023. https://www.unesco.org/gem-report/en/publication/technology
[3] Satya Nadella, public remarks on AI-assisted software development, Microsoft, 2024. https://www.cnbc.com/2025/04/29/satya-nadella-says-as-much-as-30percent-of-microsoft-code-is-written-by-ai.html
[4] Linux Foundation, Open Source Jobs Report, 2023. https://www.linuxfoundation.org/research/open-source-jobs-report-2023
[5] International Telecommunication Union (ITU), Measuring Digital Development, 2024. https://www.itu.int/itu-d/reports/statistics/facts-figures-2024/
[6] Octoverse: A new developer joins GitHub every second as AI leads TypeScript to #1. https://github.blog/news-insights/octoverse/octoverse-a-new-developer-joins-github-every-second-as-ai-leads-typescript-to-1/
[7] McKinsey Global Institute, The Economic Potential of Generative AI, 2023. https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier
👋 Hello! My mission with Beyond with Yon is to help solve humanity’s greatest existential challenges and advance the human condition. Connect with me on LinkedIn and X.