Editor’s note: Today’s guest post is by Khalid Saqr and Gareth Dyke of the Thomas Kuhn Foundation. Khalid is a fluid dynamicist, academic, and deep-tech expert. Gareth is an accomplished researcher, author, and journal manager with over 380 peer-reviewed publications.

In every generation, progress depends on one simple but radical idea: That we can understand ourselves better and act on that understanding. Nations have done it through democracy. Economies have done it through statistics. Medicine has done it through clinical trials. Science, for all its power, has not yet done it for itself, although much research has been carried out in this area.

We have built an extraordinary global system of research, millions of scientists, billions of dollars in funding, and a flood of papers each year. Yet the one thing we still cannot see clearly is how knowledge itself evolves. We can use techniques like topic modeling and knowledge graph analysis to speculate, but these look back in time and often make inaccurate, pattern-based predictions. We can measure how often something is cited, how widely it is read, or how much it is funded, but not what it means for the direction of human understanding. We have no real-time picture of where science is stable, where it is drifting, and where it is quietly preparing for revolution.

Data analytics and insights powered by big data and artificial intelligence technologies. Data mining, filtering, sorting, clustering and computing by AI for business analytics dashboard with charts.

Thomas Kuhn’s Insight

The philosopher Thomas Kuhn’s great contribution in his work The Structure of Scientific Revolutions was to show that science does not progress in a straight line. It moves through periods of normality punctuated by crises and shifts of paradigm. Those shifts are not accidents; they are built into the structure of knowledge. But because the research world has grown vast and fragmented, those moments are now almost impossible to perceive while they are happening. We see them only in hindsight, long after new ideas have reshaped the landscape. One good example is CRISPR; first published in 2012, this revolutionary gene editing approach was thought to be ‘just another prokaryotic immune system” for a number of years before its true potential was eventually realized (2015).

The Thomas Kuhn Foundation was created to solve this problem. Its mission is simple: to give science the means to see itself. We have taken Kuhn’s philosophical insight and turned it into a working system that can map how knowledge changes in real time. This system is called KGX3; it detects and classifies the epistemic function of research papers. KGX3 moves beyond keywords and citation metrics to provide a complete understanding of what a paper does: confirm, stress, or break a paradigm.

The KGX3 Engine

At first glance, the KGX3 engine looks like an ordinary digital service. It is, in technical terms, an API: a programmable service that analyzes research papers and returns structured data. But its purpose is far more ambitious. It allows any institution to ask a profound question: What is happening inside science right now?

When a research paper enters KGX3, the system reads it through a structured linguistic process derived from Kuhn’s theory of scientific development. It determines whether the work is reinforcing an existing model, identifying anomalies, or proposing something new. It assigns each paper a place in the cycle that Kuhn described: normal science, model drift, crisis, revolution, or paradigm shift. The system does this using a series of simple rules, or ‘language games’; this means that the same paper will always produce the same result. KGX3 is therefore transparent, auditable, and explainable.

The power of this approach lies not in a single result but in aggregation. When thousands of papers are analyzed together, patterns begin to appear. A field that once looked stable may show signs of tension. Another may reveal growing coherence around a new framework. KGX3 means that, for the first time, the scientific enterprise can monitor its own conceptual movement as it happens.

This capability matters because knowledge has become the decisive force in the world economy. Governments invest trillions in research, yet they still make decisions largely through reputation, inertia, and political pressure. The information they use is retrospective. A citation index tells us where attention has accumulated, not where discovery is unfolding. A grant database shows how money was spent, not what ideas are emerging.

KGX3 changes that. It provides one potential early indicator of where paradigms are forming or fading. It turns epistemic activity, the process of thinking and discovery, into structured information that can be shared. The Foundation offers this not as a product to be sold, but as a service for the public good. That choice matters. It means the infrastructure of scientific intelligence remains open, accountable, and aligned with the mission of advancing human understanding.

Collaboration Across the Research Ecosystem

In practice, this approach allows for a new kind of cooperation across the research ecosystem. Funders can identify early signs of stagnation or opportunity. Universities can align teaching and investment with emerging paradigms rather than those already passing. Publishers can see when a field begins to reorganize its own foundations and adjust accordingly. Even governments can use epistemic data to guide policy — supporting areas of conceptual renewal before they reach crisis.

The decision to deliver KGX3 as an API rather than a proprietary platform was deliberate. The Foundation’s view is that scientific intelligence should be a shared utility, not a private asset. Just as electricity powers diverse industries, KGX3 provides a common source of structured insight that others can build upon. The Foundation maintains the integrity of the system while allowing others to innovate with it. This model preserves independence, transparency, and collective benefit.

Behind the technology stands a legal and ethical framework designed to ensure public trust. The Foundation is a company limited by guarantee, with no shareholders and no distribution of profit. All income is reinvested in research, education, and public engagement, and the API is made available under defined per-use licenses. The Articles of Association lock the mission to the advancement of science and the protection of its intellectual property, which includes KGX3 and all outputs derived from it. Every corporate user must declare its applications, ensuring accountability and visibility.

This governance model may seem technical, but it serves a deeper purpose: to protect science itself from the distortions of short-term commercial pressure. The Foundation’s Fellows, Founders, and Managing Members work within a structure where every decision, every licence, and every collaboration must be consistent with the public mission. The goal is to keep the infrastructure of knowledge independent of the incentives that so often shape it elsewhere.

The effect of KGX3 is to turn the abstract idea of epistemic health into something that can be measured. We can now describe, with evidence, where a discipline is expanding, consolidating, or fragmenting. We can identify the early signs of conceptual renewal long before they appear in citation graphs or policy reports. We can, in other words, manage knowledge as intelligently as we manage economies or health systems.

What Does This Means for Science? What Will the Thomas Kuhn Foundation do?

In the twentieth century, nations built the physical and digital infrastructures that allowed commerce and communication to flourish. In the twenty-first, the task is to build the epistemic infrastructure that allows knowledge to flourish. It is no longer enough to fund science generously; we must also understand its inner dynamics. Without that understanding, we risk confusion; funding what is fashionable instead of what is transformative.

The principle is not new. Progress has always depended on measurement. When governments began measuring employment and inflation, they gained the tools to stabilize economies. When medicine learned to measure infection, it gained control over disease. When we learn to measure the evolution of ideas, we will gain control over the pace and direction of discovery.

That is the significance of the work the Foundation is doing. It is not about building another layer of bureaucracy but about equipping the scientific world with the means to make rational, timely, and fair decisions. It is about transparency where opacity once ruled (for example in the use of citation counts and ‘epistemic health’ metrics). The Thomas Kuhn Foundation has the over-riding goal to restore our confidence that knowledge itself can be understood, not just accumulated.

Of course, any system that monitors science must be used responsibly, perhaps audited by external and internal watchdogs as well as the open peer review of algorithms and open data principles. We have released a series of White Papers as well as content on arXiv and GitHub to ensure our development process is transparent. Epistemic intelligence can inform but not dictate. It can reveal where attention is needed but not replace judgment. The Foundation’s role is to provide the data, not the verdict. It gives researchers and policymakers a mirror, not a map with fixed directions. Science remains a human enterprise; KGX3 simply gives it clearer vision.

There is also a moral dimension. When the infrastructure of understanding is controlled by private interests, society loses something fundamental—the ability to deliberate collectively about its future. By holding KGX3 as a public good, the Foundation safeguards that collective capacity. It ensures that the measurement of knowledge remains independent, open, and accountable.

The long-term ambition is straightforward. Just as we now take for granted that we can see the state of the economy or the weather, we should one day be able to see the state of science. We should be able to ask, at any moment, where human understanding is converging and where it is breaking new ground. That visibility will make discovery more efficient, collaboration more intelligent, and policy more evidence-based.

What Thomas Kuhn offered in theory, the Foundation is now realizing in practice: a way for humanity to know when it is on the verge of changing its mind. That capacity is what will define the next stage of civilization.

The promise of science has always been that reason, applied with discipline, can improve the human condition. But reason itself now needs a new instrument. KGX3 aims to provide that instrument, or one route to take. Our ultimate goal is to collaborate to give science the feedback loop it has always lacked: The ability to observe its own evolution and act accordingly.

This is the quiet revolution of our time. It will not come with fanfare or ideology. It will arrive through data, through transparency, through the steady recognition that the progress of knowledge can be as measurable as the progress of anything else. It will allow us to replace guesswork with insight, delay with foresight, and fragmentation with coherence.

The Thomas Kuhn Foundation is building this future patiently, methodically, and in the open. It invites collaboration from all who share its conviction that science, like any great institution, must now develop self-awareness. For in a century defined by complexity, the greatest advantage a society can have is not merely to know more, but to understand how it knows.

Our shared mission should be to make science conscious of itself, to turn reflection into infrastructure, and to give humanity the ability to see its own progress while it unfolds. We firmly believe that such self-awareness will require a collective effort across funders, publishers, librarians, and, most importantly, researchers, to ground this vision in community collaboration.

Khalid Saqr

Khalid Saqr

Khalid Saqr is a veteran computer simulation engineer, academic, and deep tech expert. Khalid runs KNOWDYN (UK), and sits on the boards of ScienceWerx (USA) and SAQR Group (EGY). His intercultural initiatives help shape AI policymaking across MENA and bridging the gap between governance and practice. Khalid contributes to the global scientific community through his editorial roles with Springer-Nature, Frontiers, and MDPI.

Gareth Dyke

Dr. Gareth Dyke is an accomplished researcher, author, and journal manager with over 380 peer-reviewed publications. With extensive experience bridging academia and publishing, he has worked with Charlesworth, TopEdit, Edanz, and Springer Nature. Currently, he serves as Academic Director at ReviewerCredits, Sales Director at 4Evolution, and is a co-founder of Sci-Train. Holding a PhD from the University of Bristol, he has held faculty positions at University College Dublin and the University of Southampton. Gareth is also an experienced educator, delivering global researcher training sessions and collaborating with institutions across Europe and Asia.

Discussion

3 Thoughts on "Guest Post — Building an Intelligence Infrastructure for Science"

Does the tool allow analysis of theses and dissertations, and other papers that may not have a DOI?

Hi Monica- yes: you can upload any text into the working demo – https://preprintwatch.com/ – so long as the end of the file is “.pdf” … do get in touch if there are any issues … – Gareth

Yes, Kuhn certainly made us see “how knowledge itself evolves.” As a biochemist, I read it decades ago in terms of rate-limiting steps. Sometimes we need a Darwin to point us in the right direction. Sometimes, we need the collective work of regular researchers to provide the details that allow the Darwin’s to do their stuff. If, “in every generation, progress depends on one simple but radical idea,” what if we fail to appreciate it and do not grant it peer-review funding? Darwin was independently wealthy but even he needed a “bulldog,” Thomas Huxley, to overcome the opposition. And Mendel’s discovery of what we now know as genes was essentially ignored for 35 years.

A timely modern example was Katalin Kariko’s idea that led to mRNA-based vaccines. For decades, both in Hungary and the USA, her grant applications were turned down. So when the pandemic arrived the authorities could only grant provisional approval. Indeed, “Science, for all its power, has not yet done” much, and what “research has been carried out in this area” has not been sufficient to overcome the quick-fix ideas that, by their nature, are dominant in democracies.

Comments are closed.