Pod Insights The Brand Called You Redefining Knowledge Discovery | Prof Barend Mons, Professor Emeritus, Leiden University About this episode In this episode of the Brand Called You, Prof Barend Mons, Professor Emeritus at Leiden University and founding director of the Leiden Institute for Fair and Equitable Science (LIFES), shares his extensive journey in data science and bioinformatics. The conversation delves into the evolution of FAIR principles (Findable, Accessible, Interoperable, and Reusable data), the challenges of knowledge discovery at the edge of chaos, and the future of scientific communication. About Prof Barend Mons Professor Mons is the professor emeritus at the Leiden University. He's the founding director of the Leiden Institute for FAIR and Equitable Science, which is the acronym for L, I, F, E, S. In 2024, he was appointed as a Fellow of the International Science Council.
Transcript 0:08 Speaker 00 Welcome to another episode of The Brand Called You, a podcast and podcast show that brings you leadership lessons, knowledge, experience and wisdom from thousands of successful individuals from around the world. I'm your host, Ashutosh Garg, and today I'm privileged to welcome a very senior, respected and accomplished professional from the world of academia, from the Netherlands, Mr. Barend Mons. Barend, welcome to the show.
0:33 Speaker 01 Thank you. It's a pleasure.
0:35 Speaker 00 Thank you. Professor Mons is the Professor Emergence at the Leiden University. He's the founding director of the Leiden Institute for Fair and Equitable Science, which is the acronym for L I F E S. So Barend, let me start by asking you a question. Tell me about your own journey in brief.
0:56 Speaker 01 Yes, I will try to be brief. It was a long journey, of course.
1:01 Speaker 01 So I always wanted to be a veterinarian when I was a kid, but then I ended up studying biology in Leiden. And very soon I was intrigued by malaria, not so much by the disease itself per se, but by the malaria parasite as a very complicated organism to study. And that's where I spent my first 13 years after my PhD. Then I moved briefly to the European Commission and came back and found out that knowledge discovery as a process was also worth studying, but also that biology had done a lot better than most computer scientists in storing and
1:44 Speaker 01 processing information in itself. And that's what I studied for the first 15 years or so. And then I switched in 2000 to data and bioinformatics, but always with biology in my mind. And I
2:03 Speaker 01 Also found out that there is a lot of miscommunication in science, not only between languages.
2:09 Speaker 00 Uh huh.
2:10 Speaker 01 Uh I worked a lot in Africa, of course, and you have Portuguese-speaking countries, English, French, but also a lot of ambiguity in human language in general. So I found out that actually, human language is not the best way to communicate hard, fast in science. And I started the, uh, then the FAIR principles, Findable, Accessible, Interoperable and Reusable Data in two thousand and fourteen. And that sort of took the world by storm. We just passed fifteen thousand citations, I think, in that order of magnitude, sometimes ten per day on the article. Uh, that was written. And, uh, G twenty, G seven, many other political environments and increasingly also funders
2:54 Speaker 01 are requesting FAIR data now as the output. So it sort of changed the entire face of how we work with, uh, data and how we co-publish them with the articles. Articles are in fact a very outdated way of communicating science, I think. Um, and that's what I ended up doing for the last part of my career and I just retired. And my contradictory lecture was called Farewell. Um, and in fact, you could summarize it as the machine knows what we mean. So, obliged for machines, because in the quote AI world, uh, they will be our most important assistants. So I retired a few weeks ago and, um, was, uh, very well received.
3:39 Speaker 01 So now I have time for things like this.
3:42 Speaker 00 Wonderful, wonderful. Well, and wish you all the very best in the new chapter of your life. And I'm sure from all that you've told me, you're not going to be sitting and looking at retirement, I'm sure you must have got many more plans. But you know, Brenz, when I was reading about you, you've also done a lot of work in the area of creative and strategic thinking.
4:03 Speaker 00 And the next few questions are about this. What role does creative thinking play in your approach to solving complex problems in data stewardship and bio semantics?
4:17 Speaker 01 Well, yeah, you can look at, I always like to define very well what we mean by the concepts we bring together. So creative thinking is wanting in terms of new approaches to data stewardship and how to deal better with data. But I would also like to say a few words about creative thinking in general. If we consider science as the ultimate creative thinking and investigation. One of the things we have studied for years is the actual process by which we discover knowledge, as you put. Yeah. And how machines will play a role in the next phase where we use them as our assistants,
5:00 Speaker 01 knowing that the machines are stupid, frankly. I mean, they don't have a clue what they're doing. They're just discovering patterns. Now, we also discover patterns in what we previously conceived as chaos and too complex to understand. We suddenly see a pattern and that's what we call knowledge. And when it becomes established, we call it established knowledge. So we have a actual very famous graph in physics, the salt tooth. And you see that if the patterns become more and more complex, the thing that we continuously encounter today,
5:42 Speaker 01 you need more and more CPUs or intelligent units, if you want to say that, to understand everything. But if you do that, we can endlessly add intelligence. There is a straight line, you still see all the patterns. At some point, you reach the level where your intelligence cannot go any higher. And suddenly, your understanding of any more complex patterns goes asymptotically to zero. And that's the salt tooth graph that you see in many complex physiological and physical processes. And the point of it, they call it the percolation point, where you go from order to chaos and back, whatever.
6:25 Speaker 01 I have understood years ago that knowledge discovery is at the edge of chaos. And chaos actually doesn't exist, it's just order perceived by a lesser mind, that is a famous statement. So by understanding the actual process of creative thinking
6:47 Speaker 01 and how machines are not creative at all, are just coughing up patterns. But if the patterns are too complex for humans to understand, it's still useless, you know?
6:57 Speaker 00 Very interesting.
6:58 Speaker 01 So that's where we are in knowledge discovery at the moment.
7:01 Speaker 00 So, you know, Burns, you just mentioned a very fascinating phrase, which is, you said, knowledge discovery is at the end of chaos. And I was, I beg your pardon?
7:12 Speaker 01 The edge of chaos.
7:13 Speaker 00 The edge of chaos, correct, at the edge of chaos. So I wanted to, you know, for my viewers and listeners, could you explain that with an example?
7:22 Speaker 01 Yeah, if you, you look, for example, at a very complex biological process.
7:31 Speaker 01 The ultimate one is, we have, let's say, twenty thousand genes and four million pins, you know, regulatory sequences in our genome. And of course, it's a true statement that the connection between all these data points makes us human. You know, it's the template for us being human. But of course, that's true. But we have no clue how to see the complete picture of all these genes and how they connect together. Even if we have scratched the surface now for less than one percent, it's already, if you print one of these maps on the wall, people run away because it's too complex to even look at a single physiological process, let alone the whole human body.
8:15 Speaker 01 So I'm involved in a project called the Human Immunome Project. Where they want to build the AI model. I don't like AI, but okay, it's it's it's a very strange term. But let's say the model of the human immune, a very complex system. Humans are simply unable to understand the immune. The whole immune system.
8:39 Speaker 00 Okay.
8:39 Speaker 01 So what we can do is move in the direction of this complexity, step by step, by using computers. Get more and more patterns, see connections and step by step, discover actionable knowledge, as we call it. Because the fact that all genes make us human is not actionable knowledge. It's true, but we cannot do anything with it as humans. Although computers, when we tested this, can discover all these patterns. Yeah, you've done it. And even if we take the, for example, the one percent highest score between genes and diseases.
9:24 Speaker 01 In the sense that you predict that the gene might be involved in a particular disease. The computer, if you take the top one percent of hypotheses the computer comes up with. It's still a hundred and twenty thousand hypotheses.
9:39 Speaker 00 Wow.
9:40 Speaker 01 We took ten years as humans to discover six thousand or so gene disease connections, you know. And the reviewers of that article that we wanted to publish in Nature Genetics and the editor wanted it badly. Uh, the reviewers kept saying, you know, what's the point? We cannot deal with this as humans. Where's the false positive rate? Well, there is none. They're all true. They are namely just connections. Whether there is actionable knowledge there is another story. So we have to reduce the knowledge, the implicit knowledge that is out there after publishing fifty million articles.
10:19 Speaker 01 To what we can manage as humans. That's the major challenge in the current AI-dominated hive, you know? And AI doesn't exist at all there. Everything large language models cough up is a hallucination. Only when we think, hey, that hallucination makes sense to us. We think it's knowledge discovery, but it's just one of the many million hallucinations that the thing coughs up after reading trillions of full pages.
10:51 Speaker 00 Very fascinating. And Brent, how do you balance creativity with practicality? When working on large scale initiatives like GoFair?
11:03 Speaker 01 Well, that is a very clear thing. If you are a personality like myself, you know, if you look at the profile, I'm an extreme ENFP, which is the inspirer, you know? So very strong creative thinking and pretty inspiring. That's what a lot of people told me when they talked about my say goodbye lecture. But I am very bad in executing. So people like me need custodians, the other end of the spectrum. That see lions and bears on the roads all the time and slow me down, which is sometimes a painful process.
11:46 Speaker 01 But one of the things I've learned in my career, pretty early on, even when we worked on malaria. And I I've said that explicitly in my farewell lecture. Chris Johnson, my colleague, with whom we started the malaria group here in Leiden, was in the room, he is also just retired. He stayed there for forty years. He's the ultimate custodian, top scientist, science of the soul of the nature paper.
12:14 Speaker 01 But he always sees the problems first.
12:17 Speaker 00 Wow.
12:17 Speaker 01 And we were a perfect combination for fifteen years for that very reason. So, yes, there is even an official term, monson's. Instead of nonsense, you write it with an M, my second name, my family name.
12:32 Speaker 00 Yeah.
12:32 Speaker 01 So a lot of the creative thinking that I do and I see, you know, ten years ahead and, you know, I was talking about machine readable data in two thousand and five and everybody was looking at me like, you're totally crazy. But if you don't have, in the same team, custodians that keep you realistic, you fail. You're much less effective. So I've thanked my custodians very elaborately during my farewell lecture, all of them that were in the room, saying, without you guys, I wouldn't have made half of the progress that we made.
13:10 Speaker 00 Incredible. And that's, you know, very, very modest of you to have given so much credit to all your colleagues. But my next question, Barend, is what are some of the key elements of a success, of a successful strategic plan when dealing with interdisciplinary fields?
13:31 Speaker 01 Yeah, that's interesting. I always say making a plan is the easiest thing in the world. Because plans have a tendency to only fail when they get executed. So the plan is always good. That's why the big consultancy firms make so much money. Their end product is the plan. And if it goes wrong, it's your fault.
13:49 Speaker 00 Yeah.
13:49 Speaker 01 So I think that is one of the steps that you need to make. And I think making a plan, a strategic plan that is highly complex and cross disciplinary, first of all, needs this modesty. And I'm not that people think I have a lot of modesty. But modesty, I mean arrogance means you think you know what you don't know. And we call it the Dunning Kruger effect. If you're not hindered by any knowledge, you think you know everything. We have some prime examples right now in the world scene. We won't go into details. But if you really understand the issue, your own discipline, yeah,
14:34 Speaker 01 then you see how incredibly complex things are. So you underestimate the complexity of the other discipline structurally.
14:43 Speaker 01 How difficult can it be to just have more compute power? Well, it is pretty complex. Yeah. So with the cross-disciplinary research we need nowadays that goes across all disciplines, every sustainable development goal needs cross-disciplinary research and long-term strategic planning.
15:02 Speaker 00 Correct.
15:04 Speaker 01 There is two major issues. Most people do not respect the complexity of the other discipline. So you get a very difficult conversation. There's one. And secondly, the data that is underlying all these problems and the solutions to the problems, actually, if you want to have interoperability, the I in FAIR, between data from different disciplines, this challenge is an order of magnitude bigger than having interoperable data in your own discipline, like life sciences or, you know, building or environment.
15:47 Speaker 00 Right.
15:47 Speaker 01 It is already a major challenge within a discipline. But if you cross disciplines and you talk, for example, about the concept of water, a very simple thing that is there in almost every discipline, the interpretation of water is pretty different between the C data cloud, the drinking water people in the Netherlands or chemistry, chemistry saying that's H2O.
16:08 Speaker 00 Yeah.
16:09 Speaker 01 So that is one of my current foresight is to find a way to make machines understand what we call near-sameness. And near-sameness, you talk about almost the same thing, but not exactly the same thing, is one of the major challenges when we go across disciplines.
16:29 Speaker 00 Wow. What a great response, and I think I understand some of all the things there, but I probably will listen to all of this again and understand it in much more detail. Thank you. Uh, my next question is, what are some of the critical strategic challenges you foresee for the Leiden Institute for Fair and Equitable Science?
16:52 Speaker 01 Um, it could grow to death.
16:56 Speaker 01 In the sense that uh we started this very recently and let me explain a little bit more what it is. Uh, so the L of Leiden is deliberately a little bit separate in the logo. Because it is just a start in Leiden.
17:13 Speaker 00 Mm hmm.
17:14 Speaker 01 But the goal is, in fact, if you strip down all the details, and people can go to the website. Um, we want two things. We want to create as many fair data stations around the world.
17:28 Speaker 00 Mm hmm.
17:30 Speaker 01 And my last five keynotes had the title "Stop data sharing."
17:35 Speaker 00 Mm hmm.
17:36 Speaker 01 And then made people put them on the wrong leg because essentially I end by saying now we go to data visiting.
17:45 Speaker 00 Hmm.
17:45 Speaker 01 So the biggest challenge that we have, not per se as LIFES, but LIFES addresses that challenge, is that data cannot be sent around the world anymore.
17:57 Speaker 00 Hmm.
17:57 Speaker 01 In many cases, they are too sensitive. Let's take this example of the Human Immunome Project. I don't think you would argue that Chinese hospitals will send their data to the United States or the other way around.
18:10 Speaker 00 Yeah.
18:10 Speaker 01 To be analyzed. And actually, the Cloud Act is now of a major concern because if your data are on an Amazon or a Microsoft or a Google, an American-owned server, regardless of whether that server is in Europe or Africa, the US government can demand the data in exceptional circumstances.
18:30 Speaker 00 Hmm.
18:30 Speaker 01 So it is prohibited in many countries to put things in an Amazon cloud, for example. So we need to keep the data locally.
18:40 Speaker 01 Under the custodianship of the people that are custodians, in many cases even the citizens themselves, they decide how their data can be reused.
18:49 Speaker 00 Mm hmm.
18:49 Speaker 01 In addition, the data are frequently too large. If you want to send, even today, a petabyte or even exabytes of data around, it's far too expensive.
18:59 Speaker 00 Mm hmm.
19:00 Speaker 01 But if you send algorithms to the data. They are infinitely much lighter than the data themselves.
19:07 Speaker 00 Hmm.
19:08 Speaker 01 And virtually every complicated question can be answered by distributed learning.
19:13 Speaker 00 Hmm.
19:14 Speaker 01 So the data stay where they are, that's what the LEAFES is promoting. And what we have decided to be is an association, not for profit, bringing together the people that are challenging, that address these challenges.
19:29 Speaker 00 Mm hmm.
19:30 Speaker 01 So they either work on fair data, fair data stations, data stewardship, or they deliver very clever algorithms, we call them the trains or the drones that visit the data. And all of them are frequently too small or too specialized to really uh get into this market, which is an internet for machines.
19:53 Speaker 00 Mm hmm.
19:53 Speaker 01 Where you cannot demand top-down by anyone that everybody should use the same standard. So the big challenge is what we call the hourglass model in the internet. As long as you stick to TCP IP, you can build whatever you want.
20:09 Speaker 00 Hmm.
20:10 Speaker 01 And fair digital objects, the machine knows what you mean, machine actionable units of information are at the center. And LEAFES brings together everyone who believes in this principle.
20:23 Speaker 00 Mm hmm.
20:25 Speaker 01 So the challenges of the LEAFES members I just described.
20:30 Speaker 00 Hmm.
20:31 Speaker 01 The challenge of LEAFES itself so far is that we have so many people that want to join that it's for me almost a day task to talk to them, explain it over and over, then determine their role in LEAFES. They all keep doing what they're doing. It's a public-private partnership. Many of them are companies.
20:52 Speaker 00 Yeah.
20:52 Speaker 01 Many of them are too small to survive in this new, very turbulent environment, like in the beginning of the internet.
21:00 Speaker 00 Yeah.
21:00 Speaker 01 And LEAFES brings them together. That's a big challenge.
21:03 Speaker 00 Very interesting. Thank you. And my last question to you. How do you envision the future of Open Science? And what creative strategies do you believe are essential to shape it?
21:18 Speaker 01 Okay, first of all, we should get rid of the term "open".
21:22 Speaker 00 Okay.
21:22 Speaker 01 I want to be a bit provocative at the end.
21:24 Speaker 00 Okay.
21:25 Speaker 01 Uh, "open" is a term, a concept, I have really become allergic to, although I am a big promoter of open access, open science and so on. Why? There's a number of problems here. The silverback scientists that want to keep the old system going, because that made them big, always put open science as secondary, second level science. Ah, yeah, all these other science. The cells and the nature papers, that's the only real science.
21:55 Speaker 00 Hmm.
21:56 Speaker 01 So that's a downside. The other thing is, if a scientist hears "open", he or she hears "kazatui", "free".
22:06 Speaker 00 Mm-hmm.
22:06 Speaker 01 Open data is very expensive.
22:10 Speaker 00 Correct.
22:10 Speaker 01 Especially if you want to make it available for reuse in real-time setting. Ten times more expensive than just storing it somewhere on data, or whatever.
22:20 Speaker 00 Yep.
22:21 Speaker 01 So there is a lot of money involved. About 5% of all research should go into this data stewardship and provision of data.
22:27 Speaker 00 Mm-hmm.
22:28 Speaker 01 But also, a lot of data cannot be open. So if people say "open access", yeah, of course, articles should be open access wherever possible, because it's not sensitive data. But I'll give you two examples. I am in the Leiden University Medical Center. There's no way we can make our patient data open, of course.
22:50 Speaker 00 Yeah.
22:51 Speaker 01 They can still be fair and visitable under well-defined conditions, especially if we don't give out the data, we just answer questions in the hospital for even the pharma, no problem. Another example is a sibling institute that we want to start in India. We want LIVS-type institutes around the world. The Indian one is working on hidden cameras that snap tigers.
23:16 Speaker 00 Hmm.
23:17 Speaker 01 Imagine if you publish those data in open access, the tigers will be dead the same afternoon by publishers.
23:24 Speaker 00 Yeah.
23:25 Speaker 01 So "open" is not that all the data are open, all the software should be open source. Open science to me means science done right.
23:34 Speaker 00 Mm-hmm.
23:35 Speaker 01 And the science of the past decades, where everything is behind firewalls and there is this ivory tower group, that should be stopped. And we should really go to fair science. That's why we call it the Leiden Institute of Fair and Equitable Science and not Open and Equitable Science, okay? Equitable access to fair data, many of the data are highly sensitive, will not be open. So, open science is, in that sense, its own worth envy. Because if you say everything should be open, the connotation is, it's free.
24:15 Speaker 00 Right.
24:16 Speaker 01 Scientists think they can get these data for free because everything is open. No, open science is needing five percent of all expenditures on science. And our cabinet just put seven billion less, they are absolutely crazy, but okay, that's what the right wing government is doing. It's that five percent of that investment should go in making data fair and exposing it in responsible way for others to reuse around the world. And there should be equitable reuse, so it should be as easy for an African scientist with one iPad.
25:00 Speaker 01 As for Professor Leiden or Harvick. That's what we mean by equitable.
25:05 Speaker 00 Amazing. And on that note, Professor Mons, thank you so much for speaking to me about your journey. I think I learned many things from you about the Leiden Institute, Leiden Institute for Fair and Equitable Science. I also learned a lot of things from you about data, about not open science but fair science, not open data but fair data. And of course, I love the way you told me that knowledge discovery is at the edge of chaos. Thank you so much for speaking to me, sir, and all the very best to you.
25:41 Speaker 01 It was a pleasure.
25:45 Speaker 00 Thank you for listening to The Brand Called You, videocast and podcast. A platform that brings you knowledge, experience and wisdom of hundreds of successful individuals from around the world. Do visit our website, www.tbcy.in, to watch and listen to the stories of many more individuals. You can also follow us on YouTube, Facebook, Instagram and Twitter. Just search for The Brand Called You.
Previous Changing Landscape of Investment Banking and Corporate Law | Rosalyn Breedy, Managing Director, Breedy Henderson Ltd Next Evolution of EdTech | Alexander Zheltov, CEO & Co-Founder, Educate Online Inc Page updated November 15, 2026 · transcript and summary are machine-generated