Kuhn, AI, and the Changing Paradigm of Learning 

Kuhn, AI, and the Changing Paradigm of Learning 

For the last couple of weeks, I have been reading a book titled The Structure of Scientific Revolutions by Thomas S. Kuhn. I came across a very interesting argument in it—one that deeply captured my attention:

“So long as the tools a paradigm supplies continue to prove capable of solving the problems it defines, science moves fastest and penetrates most deeply through confident employment of those tools. The reason is clear. As in manufacture so in science—retooling is an extravagance to be reserved for the occasion that demands it. The significance of crises is the indication they provide that an occasion for retooling has arrived.”

I have been thinking about the kind of crisis that we are currently witnessing in education. Across the world, there is growing concern about students’ learning, their ability to comprehend, and the extent to which our existing systems are preparing them for the world they are going to inherit. I believe this is where Kuhn’s argument becomes particularly relevant. The significance of a crisis lies in the indication it provides that an occasion for retooling has arrived.

What could this retooling be?

We will have to look at how our learning needs change with shifts in paradigms. The Scientific Revolution and the Industrial Revolution were, in their own ways, fundamental paradigm shifts. They changed not merely the tools available to human beings, but also the ways in which we understood the world, organised our societies, and defined progress. Much of modern schooling, as we experience it today, is tied to the industrial paradigm. Age-based classrooms, standardised curricula, fixed timetables, uniform assessments, and the idea that knowledge can be delivered to a large group of learners in a predetermined sequence all bear the imprint of this history. The system was designed, at least in part, to organise learning at scale and prepare people for an industrial society.

But the world has changed. The nature of work is changing, the ways in which knowledge is produced and accessed are changing, and the relationship between individuals and institutions is changing. Yet, in many ways, our systems of education continue to operate with the assumptions of an earlier era. We see this crisis manifest not only in education, but also across employment, governance, social structures, and beyond. The tools that once served us well may no longer be sufficient to address the problems that confront us today. And perhaps this is precisely what Kuhn meant when he suggested that a crisis signals the need for retooling. It also signals the possibility that we are moving towards an altogether new paradigm.

In this emerging paradigm, I want to draw attention to the kind of intervention that artificial intelligence tools are making. I visualise a future where the success of a person or an institution will be increasingly tied to—and perhaps partly defined by—the extent to which they learn to leverage these tools. Take, for instance, two students, A and B, studying in school. Until recently, we might have explained the difference in their performance largely through the quality of their schooling, their teachers, their individual effort, and the sociocultural capital available to them. These factors will continue to matter. However, they may no longer be sufficient to explain the differences in the opportunities available to them.

The additional defining characteristic could be their access to AI tools and, more importantly, their ability to use them meaningfully. A student who knows how to ask the right questions, evaluate the responses generated by AI, identify inaccuracies, connect ideas, and use the tool to deepen understanding may have a considerable advantage over another student who does not possess these capabilities. This is not to suggest that access to AI will automatically translate into better learning. An AI tool can just as easily become a shortcut that discourages independent thinking. The difference will lie in how it is used, the quality of guidance available to the learner, and the capacity to question rather than blindly accept what the tool produces.

The same holds true for institutions. An institution’s performance will no longer depend solely on the calibre of the personnel it possesses or the conventional resources it can access. While these remain baseline conditions, its success will increasingly depend on how well it develops coherent policies to leverage AI tools. This includes preparing its people, rethinking its processes, safeguarding data, addressing ethical concerns, and ensuring that technology strengthens rather than replaces human judgement.

For schools, this could mean rethinking the role of the teacher, the nature of classroom interactions, the design of assessments, and even the very meaning of learning. If information is increasingly available at our fingertips, what should children learn? If AI can generate an essay, solve a mathematical problem, or summarise a text, how do we assess whether a child has actually understood it? If a machine can provide answers, perhaps our greater responsibility is to help children develop the capacity to ask better questions. And what happens to those who do not have access to these tools, or who are not taught how to use them? Could AI create a new divide between those who can leverage it and those who cannot? The digital divide may no longer remain limited to access to devices and the internet. It may increasingly become a divide in the ability to use technology to learn, create, and participate in society.

All of this brings us back to the original question: could the adoption of AI be the very retooling Kuhn spoke of?

I believe it could be a part of it. But retooling must mean much more than introducing AI tools into classrooms or asking teachers and students to use new applications. If the underlying assumptions of schooling remain unchanged, we may simply end up using new tools to reproduce old practices. We might automate the very processes that need to be questioned.

Kuhn further writes:

“Sometimes the shape of the new paradigm is foreshadowed in the structure that extra-ordinary research has given to the anomaly.”

I find this thought particularly significant. Perhaps the anomalies we are observing today—the persistent gaps in learning, the changing nature of knowledge, the limitations of conventional assessment, and the possibilities opened up by AI—are already giving us glimpses of the structure of a new paradigm. The question is whether we are willing to recognise these anomalies for what they are. Are we willing to question the assumptions that have shaped our institutions for generations? Or will we continue to defend the existing paradigm simply because it is familiar to us? I do not think we yet know what the new paradigm will look like. But I suspect that it will redefine everything from how we learn and how we school to how we work, how we govern, and how we understand our place in society. Perhaps the real crisis is not that our existing tools are failing us. It is that we continue to expect them to solve problems for which they were never designed. And perhaps the real retooling must begin with our willingness to question the paradigm itself.