Projets Libres x Amel Charleux, associate professor at the University of Montpellier 3

[The Fundamentals] Open source as seen by a researcher – Amel Charleux

Projets Libres

–

Open Source as seen by a researcher – Amel Charleux

Walid: welcome to all of you for this new episode of Projets Libres. We are resuming a series that we started a long time ago on the economic models of free software. Today, we’re going to look at this through the prism of research. You’ll see, it’s going to be quite interesting and certainly very complementary to what we’ve been able to do before. If you want to know more before starting this episode, I invite you to go see episode 9 of season 2, in which, with Raphaël Semeteys, my friend who is here today, and Goneri Le Bouder, we made an introduction to the economic model of free software. It could be interesting, if by chance, there are terms that we are talking about today that we don’t explain directly, you can refer to this episode.

So today, I’m delighted to have with me Amel Charleux, who is a associate professor at the University of Montpellier. Amel was mentioned in an episode, which was episode 10 of season 3, where we talked about the creation of the OpenRail foundation and the OSRD software: Open Source Railway Designer. It was Loïc Hamelin who introduced Amel. And so, I took the liberty of contacting her because I was interested in having her point of view as a researcher on all the themes that Raphaël and I address.

And so, Amel, listen, welcome to the podcast Projets Libres. Thank you for having responded positively to our invitation.

Amel: thank you for the invitation.

Walid: So, for this episode, as usual, I’m with my friend Raphaël Semeteys who is also passionate about everything that revolves around economic models and the governance of free software. Raphaël, welcome to you, I hope you’re doing well.

Raphaël: It’s okay, it’s okay. Super happy to be there.

Presentation of Amel Charleux

Walid: Perfect. Well, let’s get right into the thread of the subject. For those who don’t know you, Amel, could you start by introducing yourself and giving us a little bit of your background? What did you do and how did you end up as a researcher at the University of Montpellier?

Amel: Well, hello everyone. Thank you again to both of you for this invitation. The podcast is not necessarily an exercise that we are very used to as researchers, but it is an interesting format where we can actually present a certain number of things. As you have well explained, I am a lecturer at the University of Montpellier. I am part of the MRM laboratory, so Montpellier Research in Management. My background, quickly: I did a master’s degree at the same university a number of years ago. And then, I went to work, I did internships, a number of internships. And one of them, I made it in what was called at the time an SS2I, so a computer engineering service company, which was Kaliop, I think some people know.

And so, they had a few projects that they deployed using tools, open source software. And so, it’s true that it was my first encounter with open source, this internship. Then it happened that I met my husband during this internship, who is a computer engineer, who works a lot on open source, so Hugues Charleux. And so, along the way, these are subjects on which we were talking a lot. And this work on open source, or at least this internship, has raised questions. That’s it, it raised questions because what we were taught at school wasn’t quite what I saw in that internship.

It’s true that open code, an open product, a community, collaboration, I didn’t understand. I didn’t have, in any case, the right lenses or glasses for that. I went to work for a few years and then it stayed a little bit in the background, it turned in the background in my head, these questions. And at some point, I said to myself “I have to dig more”. And so I went back to knocking on the door of the university by proposing this subject. Let’s say that I’ve always loved school, learning, and I wanted to really dig into this theme. My thesis supervisor agreed to supervise me at the time, it was Anne Mione. We worked on this subject of open source, which was, at least at the level of the Montpellier Research in Management laboratory, quite new. Then I defended and finished my thesis. I did my doctorate and then I took the exam and I was recruited here at the same university. That’s a bit for the course.

Walid: But I think it’s a subject that, overall, is not necessarily covered a lot, actually.

Amel: Not much. You’re right, Walid.

So, there are researchers who work on these themes, but we are a bit original in the laboratories. Let’s say that often, it’s more about what we call research field. That is to say, we will be interested in innovation, for example, and then we will say to ourselves “let’s see how they innovate in open source communities”. But the research theme is rather more broadly innovation.

Amel Charleux

Or we will work on the accounting and then we will look at how it is done in this or that field. And so in the end, open source is more of a field of research where we go from time to time to see how certain things are done, more than really a concept or a research theme. It was quite a long discussion within the research group to which I belonged: was open source more in my field of research or was it the concept I was really working on? And so I really wanted to work on the concept. It didn’t stop at a field of research, we call it the subject of my research.

The materials that serve as the basis for Amel’s research

Walid: And precisely, if we talk about the subject of your research, then your research is about open source. Raphaël and I were wondering about the way you worked. What materials do you use in your research?

Amel: So, for management research, in any case, we have several ways of proceeding, we have several methods. We can do what we call quantitative: we will look for databases, or we will collect responses via questionnaires that we will then process in a very quantitative way. But we also have what is being done a lot, and especially in France, the Americans and the Anglo-Saxons tend to do quantitative.

In France, we like what we call qualitative research, so research that is more contextualized, that is based on a better understanding of the environment. And we’re going to work more on interviews. We will increase the number of interviews, exchanges with people in the field, and therefore professionals. This is called primary data. We will collect them directly from the people we are interested in. And then we have everything we call secondary data, which we’re going to cross-reference with this primary data. And so this secondary data will be everything we find on the Internet, the sites, the financial reports, the minutes, all the documents that we can have access to in companies if we ask for it and they agree to give them to us, etc. So we’re going to run into each other, at least personally, I do mixing. I’ve also done quantitative data, but I’m going to be more inclined to go for qualitative and contextual data, and therefore to collect interviews and primary data, which I’ll also cross-reference with secondary data, to then have an understanding of the phenomena, let’s say. And so this understanding, it’s very contextualized. We are always very careful to explain well in our papers. I don’t know if you’ve read one or two, but we always try to explain that it’s very contextualized, that we can’t generalize our results because we’re not looking for absolute truth, but about very contextual understanding.

Raphaël: I have a question about the scope of your initial research, and then even current, during your thesis, because open source is quite global. So there, you are attached to a unit that takes care of management. So, the scope of your research, is it general like that or is it going to be with a specific angle that would be related to that, to management?

Amel: So, in the laboratory, we are organized by themes, let’s say. There is the accounting group, there is the HR group, human resources, there is the information system group, there is the management control group, and then there is the strategy group, and finally there are several groups, entrepreneurship. And so, we are part of certain groups according to our appetites. I, for example, am part of the organization group, where we will study how structures are organized. So, it’s quite broad. And under that umbrella, personally, what I’m going to look at, what I’ve been able to look at in the past, is going to be, for example, business models and the governance of open source projects. These are questions that are linked to themes of sustainability, for example. And then, what interested me a lot, I think that this too, we’ll talk about it maybe a little later, is interorganizational relations: it’s also a theme that I’m going to dig into from the angle of open source to try to understand how open source transforms, changes interorganizational or interpersonal relations as well. So these are the themes on which I have an appetite and on which I work. But I could very well perhaps, I don’t know, as my career progresses, I could move towards the study of accounting, in projects or other themes, it’s not closed. You can really have careers where you evolve, you change the theme depending on the encounters, on what comes up. It’s true that it’s quite an interesting freedom.

Raphaël: Thank you. I think we’re going to talk about this aspect again, a little bit of interaction, especially relationship. It’s not an angle that is often put forward by open source, so I’m very happy to have you here.

Walid: You can’t have Raphaël’s smile (laughs). There, you study open source, but with a rather French-speaking focus. Or does your research also focus on communities, tools, ways of doing things that would be more English-speaking or from other countries outside the French-speaking world?

Amel: Let’s say that at the beginning, to understand a little bit, to explain what open source is, the phenomenon, where it comes from, etc., we can’t stay only on France or the French-speaking world. So, we are forced to open up and look at what is emerging across the Atlantic, what the trends are, etc. So, I’ve done a lot of work, especially the work I’ve done on business models, it’s a work that I’ve done in a very cross-functional way. There were no borders, let’s say. Then, at other times, indeed, it can happen that we focus on a particular case. And there, we will look either in particular at what is happening in France or in particular at what is happening in a community or a project. And so there, we worked on a Belgian project, for example, which we studied in more depth. It was an interest of the researcher, but it was also an opportunity insofar as we had access to this field. So we were able to go and do a lot of interviews, conduct a lot of interviews and have access to a lot of documentation and data. So there you have it, we worked on this case. But there is not, let’s say, a firm desire to stay on France or on French or French-speaking projects.

Where does Amel publish her research?

Walid: Where do you publish your research?

Amel: So, on the publications, we will have several axes. It’s true that we, as researchers, are rather encouraged to publish in scientific journals. So, I publish in scientific journals, especially French-language ones, when I write in French. So, you’re going to have the SIM review, the Innovation review, Terminal. In any case, there are a few French-speaking journals that are interested in open source, because it is not necessarily easy to publish on this theme in any journal. So, we also target the scientific journals in which we are going to publish. We are going to enter into discussions with communities that are interested in these subjects, and therefore that represent themselves in these magazines. At the European level, I may publish in the European Management Journal, where we recently had a publication.

Depending on the language we choose, the case we choose, we will target more or less international scientific journals. And so that’s the priority. That’s what you’re evaluated on as a researcher: on this scientific publication in journals. So more and more, we’re also interested in how we’re going to popularize our research, how to disseminate it: it’s a fairly important underlying trend in universities and laboratories, where we’re trying to break down silos, to do a little bit of open source, because in the end, what we publish in journals, well, In any case, it is rarely consulted or rarely read by the professionals themselves. And so, it doesn’t percolate much into the field. And what we want is to have more impact as researchers and for our research to have a real usefulness, even beyond the simple construction of knowledge, because we build knowledge, it’s very interesting.

Amel Charleux

So intellectually, it’s very stimulating, but we’re encouraged and I think it’s a motivation that most researchers share to go a step further and spread more. And so, it’s true that the podcast or round table formats that I’ll be able to participate in, the thematic sessions that I’ll be able to organize, these are things that are encouraged and that I’m doing more and more.

Walid: It’s true that in other fields that are not computer science, there are quite strong relations between researchers and industry. I was asking myself the question, because in the field of open source, at least in French-speaking France, I’m not sure it’s still very developed. However, indeed, it’s something that we try to put forward also at the podcast level, it’s something that interests us a lot to have a different point of view. We experience it more as professionals, so it’s something that we experience on a daily basis in one way or another, that we have experienced, but you have an angle where you can take a step back and look and have an angle that is certainly very different from another.

Proximity to industry

Walid: Precisely, I was going to ask you how you could also popularize round tables, as I will put links in the transcript, but the round tables that you have been able to do at trade shows like Open Source Experience, which are also a good way to appear with the industry as well. I don’t know, it’s something you do regularly? Are we contacting you to do this? How does it work?

Amel: So, Open Source Experience, it’s true that I’ve done it several times and then it’s also been a very important database for me to collect data. That is to say, at the very beginning of my thesis, when I had to go and collect data, do interviews, that’s when I went to look for my interlocutors. And it’s true that the advantage with Open Source Experience, and finally the open source ecosystem, is that people are quite happy to talk about their work, what they do and how they do it. So collecting the data was quite easy, which is not necessarily the case in all industries. So that was rather a positive thing. Open Source Experience was a bit of a resource for me initially. And then I think that… I met, I discussed, I exchanged, I interviewed, and finally relationships are created, bonds are forged, and I have come back regularly to the Open Source Experience. I think it’s a fairly classic event for the French-speaking ecosystem, at least on the business side. To come back to your questioning and questioning of the link between research and the industrial and professional world, it is true that initially, in management, I am not necessarily going to speak for the other disciplines. I’m not going to talk about medicine or biology or computer science, but really about management. Initially, to be a management teacher or management researcher, you had to have some experience. And it was even necessary, it was rather valued, to be in contact with the field, to have missions. I was listening to a podcast by Jacques Igalens, who is a very, very important researcher in management, who explained that at the beginning, he did a lot of consulting and that he had three missions: research, teaching and this very field professional activity. And finally, it’s true that there has been a kind of shift and today: a teacher-researcher will do research, teaching and then a lot of administration. The contact with the field, work with the field, at least as a professional, has given way to more institutional missions. I, for example, am in charge of a degree, I am in charge of international relations within my component. So, missions that I call institutional, not so much administrative, but which, as a result, leave little room, in the end, to really be in touch with the field. So, this shift, in the end, it cut us off a little from practitioners as management researchers. And so, the only times we are in contact with the practitioners are when we go to observe them, look at them, talk to them. But we are not in action with them.

We are very little in action with them. But that’s very French and French-speaking, though. The Anglo-Saxons, on the other hand, the business side as researchers, is very unbridled, it’s very accepted. All the researchers do consulting on the side. There is no ethical problem, let’s say, in relation to that. So, I won’t take a position on this issue, but it’s true that there are also issues of independence that can be important when you go to do consulting and get paid to do things, while working at the university. It’s true that there are questions of academic freedom, especially on issues or subjects that are a little more sensitive, in pharmacy, things like that. In open source, in management, it is perhaps less serious.

Amel Charleux

And so in the end, this link between researchers and professionals, in the end, it hasn’t been broken a bit, but in any case, it is only maintained by the research side, and not the practical side. And so, it’s true that the fact that I can go to the Open Source Experience, even outside of the interview side specific to a research theme, the fact that I go there regularly, it allows me to take the temperature a little bit of the topics that are interesting for the community, for the ecosystem, what are the problems of practitioners, what they are up against, etc. So, in the end, this presence nourishes me enormously, even beyond the publication of an article. It’s really about being aware of what practitioners are going through.

The courses given by Amel at the university

Walid: Precisely, to finish on this part, you give courses that are related to your research. You have classes of students that you supervise and who work on these subjects with you, right?

Amel: I don’t have students who work directly, exclusively on open source, because I don’t have a degree in open source community management or open source projects, which might be quite interesting to put together. I have more students who are there and who do management. So, they’re going to do HR, marketing, accounting, strategy, organizational theory, things like that. I work on the theory of organizations, for example, which is something very theoretical, classical, but I will also try to bring elements related to my research, always. I always introduce myself and my research to them. And then I’m going to try to have sessions, zooms, let’s say, on how open source works. We’re going to present the classic things. And then I’m going to try to bring an alternative model, if we’re going to talk about intellectual property. I’m going to talk about copyleft, even though it’s something that isn’t necessarily covered elsewhere in traditional courses. So, I’m always going to try to bring in case studies as well. Rather than doing a case study on, I don’t know, Sony, Samsung, Ikea or Renault, I’m going to do a case study on an open source company, a project, I don’t know, we’re going to take Red Hat, we’re going to do a case study on Red Hat, on its business model or things like that. So it’s true that I bring open source and my research themes in small touches, let’s say, throughout the students’ career. And afterwards, I can also intervene, from time to time, I can be called upon for a particular course, a seminar in particular, where I will talk exclusively about open source for three hours or four hours or five hours.

Coopetition between researchers

Raphaël: Yes, I have a question about the community, but this time researchers actually. How are relations going, especially in Europe, but also perhaps with colleagues across the Atlantic? Is there competition, coopetition? How does that work?

Walid: he placed “coopetition”, that’s it! (laughs)

Amel: that’s it! Well done! So yes, clearly, there is coopetition. But in fact, coopetition, it’s true that as a research concept, it’s something quite recent.

Walid: Let’s perhaps remember what coopetition is for listeners who haven’t listened to the previous episodes.

Amel: yes,

So coopetition is a concept that reflects the fact of being in cooperation with a competitor. So to go and cooperate with someone I’m supposed to eliminate. Theoretically, in classical organizational theory, and in particular Porter’s highly competitive perspective, I am supposed to eliminate my competitor in order to have a competitive advantage and establish my position and dominate the market. Coopetition qualifies this, and therefore allows us to have a slightly alternative vision of relationships and to say that in the end, it’s not all black and white. We are not just friends or enemies, but depending on the circumstances, the needs, the situations, we may need to collaborate with a competitor at a given time.

Amel Charleux

So coopetition is going to be about how we do this, how to do it well. What rules should be respected, precisely, so that there is no opportunism, so that it does not turn against the one who is doing coopetition. So, coopetition, indeed, a research theme that I have worked on a lot and that I found very interesting to work on in open source communities, precisely. But I may come back to that later.

I’ll answer your question, Raphaël, about communities of researchers: you can either be a solitary researcher, and work a lot alone. No one is going to come and tell us, ‘no, no, you absolutely have to collaborate or work as a team’. It’s true that we have a lot of freedom. We are a bit of entrepreneurs and we run our boat a bit as we wish. We make the career, we build the career we want to have. So I tend to prefer to work with colleagues and many of us do that. So, we will develop collaborations that often emerge at conferences. So, it’s kind of our own Open Source Experience. So, we have several by theme, precisely: there are conferences specialized in accounting, others on HR, others in IS. There are some that are a little more transversal, management conferences, more general. And so there, we’re going to often present our work, what we’re working on, so what we call communications, we’re going to communicate. And then, we have rooms where we have colleagues. And that’s how connections are made. We say to ourselves “Ah, but wait, I’m working on this too. Look, maybe we can move forward together.” And that’s how we collaborate.

That’s it, we collaborate and the scientific community is created. It’s really around these conferences and these communications. We don’t really have a particular online space where we can go and propose topics or a GitHub for researchers, for example. We don’t really have that. Or we’re going to launch an idea, or start an article, and then someone would come and graft on, complete, increment, etc. We don’t really have that. So, that’s why there’s a bit of competition in the sense that we still work in small groups or in silos. So, it’s always a bit benevolent, but a bit of competition on certain subjects. When there are too many people on certain subjects, we try to be the one we see the most or that we will quote the most or we will try to differentiate ourselves, let’s say, to have a little differentiation despite everything. So, there are a bit of the mechanics that we find in companies too, of differentiation in this ecosystem.

Walid: In our ecosystem, we have a very important show called FOSDEM, in which there are discussions. It’s more of a developer-based show, but in which there are discussions that aren’t developer discussions. I am thinking in particular of everything related to the law, European policies, etc. And so, in which there are very interesting discussions and which are also a place where people who work on these subjects meet. I was wondering if these are places where you can go, find yourself? Have you already, for example, been to this salon, which is in Brussels, which is usually always the first weekend of February?

Amel: So, FOSDEM, I’ve been saying to myself for years, well, come on, this year, I’m doing it. Come on, this year, I’m doing it. We also have time constraints to be very practical, I want to say. So, it’s true that it’s a show for developers. But these are things that interest me too and I think I could really feed off a lot of things in this living room. It would be very interesting in relation to my research. But at some point, I also have to make very practical choices about time. That is to say, if I am going to do two research conferences plus one or two round tables, it will already take me a lot of time. And then there are also questions of financing since all these trips, all these trade fairs… These conferences must be funded. And we don’t necessarily have unlimited budgets in our public laboratories. We have to make choices. But I will. It’s something I’ve been thinking about for some time. But one day, I will.

Walid: I’ve been going there almost every year for a long time. Luckily for us, the lectures are recorded and broadcast. And this is, I must admit, my main subject to work on the podcast, the FOSDEM conferences.

Amel: It’s true that that’s interesting. But it’s more of a “we take” business. And it’s true that we don’t necessarily create connections.

Walid: No, you have to be there physically.

Amel: Well, you have to be there physically to create the conversation, create the exchange and be able to dig into some of them.

Raphaël: And God knows that at FOSDEM, that exists there.

Amel: the connections?

Raphaël: Yes, there are a lot of people. It’s very dynamic. You meet a lot of people there, it’s very interesting.

Amel: Loïc goes there often.

Walid: yes, Loïc Hamelin from OSRD and OpenRail and I, I met people I had interviewed in person or I got in touch with people I wanted to interview at FOSDEM. The best example is… there are conference rooms, for example the Social Web conference room, so that’s where I’m going to come up with ideas. For example, I met the people of Mobilizon, whom I interviewed (here), I talked to them there, I introduced myself and everything. It’s a pretty unique opportunity to meet people and talk to them. If you’re not shy, you can take someone, ask them for five minutes. It’s really great. I talk about it on every episode, I think, of FOSDEM.

Collaboration with industry: the case of OpenRail

Which is related to what we have just said. I discovered your work through the collaboration you made around the creation of the OpenRail Foundation. So, foundation, business, which aims to share software created by infrastructure managers in European countries. And now, it even comes out of European countries, since there is also the ONCF, Morocco, which has joined in this. So, one of the questions I asked myself about this was your relationship with the industry. Was this collaboration with the SNCF something one-off? How did it go? And then, the next question, of course, is what did you, as a researcher, get out of this collaboration? How did you participate? What did you get out of this collaboration?

Amel: Yes, so it’s a collaboration that we started at the very end of my thesis. In any case, an intern that Loïc had recruited to work on this issue, had come across one of my articles on business models. And so, she had him read it and they were like, “Ah, but that’s exactly what we’re working on.” And so, they contacted me and then we started with one meeting, then two, and then it kind of set up like that. They presented me with the project they had, and then the help they needed, let’s say, to be able to move forward. And so it started anyway from this research article. I was saying earlier that in the end practitioners have difficulty accessing these articles because it’s always a bit long, there is a theoretical case, there is the methodology, sometimes before getting the result, you have to hang on. That’s why we try to make versions a little lighter in media that are a little different from scientific journals, but sometimes it’s read. And that’s an example of someone in the field, a practitioner who immersed himself in a scientific article and found that there was interest. And so then, I accompanied them in the reflection, in the structuring of the reflection a little bit, to bring out, let’s say, the questions that need to be asked and then perhaps also to pay a little attention to the blind spots. That’s really what work was all about. To try to show what the possibilities were when there were several choices to be made or when there was at a given moment an impasse, what were the possible solutions. And they made their choices, they moved in the direction they wanted, which seemed to them to be the most relevant for them.

Loïc is a really exceptional person, he’s very curious, he really manages to try to be very empathetic about the different models, the different situations, and then to project himself a little bit. He really manages to have this long-term vision. And so, he managed, I think, to carry out this project that he initiated from this foundation. And then there was a whole team around. And so, it was really working with them, a lot of meetings, and then proposing, showing the different options each time possible, and they, they chose the directions they wanted to take. So, it’s true that this collaboration has been very rich. It came at the end of my thesis, so it’s true that I had already worked a lot on other fields, on other materials. That was a particular case, while I haven’t yet exploited it, in terms of research, that is to say that I haven’t yet published on this case. But it’s true that I’ve all told them, they’re all aware that as soon as I have a few moments, I think I’ll contact them one by one to do interviews. And then, maybe it’s not bad to take a step back, that it also settles on their side, that it has had time to settle. And that we can do a series of interviews and then perhaps consider a scientific publication on this particular case.

Cases like this, which are very industrial, allow you to really be at the heart of the mechanics and not just to observe it from the outside as a researcher and therefore to perhaps miss certain things: negotiations, for example, on this or that option, to see the different positions, the different perspectives according to the roles of each person. So all this is very interesting when you are really at the heart of things. And then, obviously, in the long term, as a researcher, the objective is to advance knowledge in the end and for it to be useful to others. If others, tomorrow, wish to launch a foundation, well, if I publish on it, there will doubtless, I hope, be useful.

Raphaël: I think it’s great, this way of finally collaborating with practitioners, as you said, with the industry. It’s a win-win for everyone, because the researchers, you are at the heart of the action, of the practice. As you said, you can see things that are not necessarily visible otherwise. And for the practitioners themselves, it’s benefiting from this hindsight and this somewhat global vision, you were talking about blind spots, etc. Sometimes, when you’re caught up in the heat of the moment, you have a hard time having that posture. It brings a lot of value to everyone to do this kind of thing.

Amel: I had a kind of neutral position, which means that I didn’t have an interest in it going one way or another, that there was a decision that was made or another. So, it’s true that this somewhat neutral position can help when you’re setting up projects and when, sometimes, you ask yourself questions and you don’t know where to go. Seeing a somewhat neutral vision can indeed… And I think we don’t know this enough, but researchers are still very open to this type of thing. They are rather demanding and we don’t think about it. I was listening earlier… No, it wasn’t just now, it was… last week, a podcast you did with Sébastien Dinot.

Walid: Yes, absolutely.

Amel: on governance. And so, at some point, it is a question of who can help in the reflection around governance. And so, he did indeed cite a lot of possible interlocutors, but it’s true that researchers have not appeared, for example, in this landscape, people who can help think about governance or an economic model or about questions of organization, structuring. It’s true that we don’t think about it as a practitioner, but it’s also a little bit our fault. So, we need to be more able to reach out to each other.

Raphaël: The fact that you’re in the podcast really goes in that direction. That’s great.

Walid: It’s a subject that, personally, interests me quite a bit because I listen to other podcasts in other universes in which there are a lot of researchers involved. And what’s interesting is precisely because they come up with a methodology, a different way of seeing things, a different language too. And in fact, when you listen to that, it allows you to take a step back. And so, that’s why it was also one of the big goals that Raphaël and I had for the future, was to have more researchers come to talk on the podcast, because it allows us to broaden our horizons a little bit already on our own, to take a step back from that. And so, there you have it, this is something that’s super interesting. And normally also, on these foundation topics, well to set up a foundation, etc., we should have other quite interesting things coming up on the podcast in the coming months. It’s under discussion, but it takes time. But we should have some really cool topics coming up too, a bit in the same vein as OpenRail.

Sovereignty and Open Source

Walid: Before moving on to your point of view and your view on open source, the last thing we asked ourselves and which is very topical, so it is a bit like the OpenRail part, where the goal is to say to ourselves “we all develop software, and then at the European level, we can share and learn, etc”, this is the notion of sovereignty that is very fashionable now. And so, when we talk about sovereignty, we often talk about open source. If you want to say a few words about this, how do you, as a researcher, see things? Is this something you follow? Does this seem interesting to you?

Amel: So, it’s something that I am and that I live. That’s it, let’s put it like that.

On the question of sovereignty, I think it’s a hot topic, let’s say. Right now, with everything that’s going on in the United States, Donald Trump, I don’t see who could say that it’s not an important subject today. Sovereignty, in any case, not as something that would lock us up and cut us off from others, is not in that sense. That doesn’t prevent us from working with American publishers, with Chinese, Asian or whatever. The idea is not to close ourselves off and withdraw into ourselves, but rather to guarantee a certain independence and a certain capacity to secure these information systems. That’s really it. And so I think this question can only be addressed through open source. I’ve turned the question around a bit and I think that really, the only possible way out is the open source one.

Amel Charleux

And we need to address this issue in any case, but not on an individual level. So it’s important that at our level, each of us as a user, we are aware of this and that we pay attention to it, and that we perhaps promote more European solutions, I’m not going to say national because that’s not the objective either, but already European. Nevertheless, I think that there is something that is at stake at a more meta level and rather at the level of the institutions that govern us.

And when I say that I am living this situation, I will tell you a little anecdote. I’m in a management department and so there are several teacher-researchers and then we tried to modernize our work processes and therefore avoid sending emails, files, updating students, things like that, trombinoscopes, schedules. And so we said to ourselves we are going to use a solution and so Teams came on the table because we had Teams in our ENT, in our workspace, we have the whole Office suite and then we have Teams which is made available. So I’m pushing not for Teams, but for more modernization of our work processes to make everyone’s life a little easier. And then, I still have this concern about the issues of protecting our students’ data, our own data, and the use of Teams and Microsoft in a public institution, in a public university. And then finally, while talking really by chance in a corridor with a colleague, he told me but we at the CNRS, we don’t have the right to use Teams, if in fact we use Resana. I say, but what is this, Resana? It’s a bit like Teams for public service employees. I say, how? Excuse me? But I’ve never heard of it. And so, things are being done. There are plenty of solutions. And so, I saw that we could have this space for collaboration, manage projects, that there are spreadsheets, word processors. Really, it’s rich, it’s great, Resana. But I discovered him in a hallway, while chatting randomly with a colleague. And so, that’s where I say that there is perhaps a more meta issue where, at the institutional level, there should be more pushy people on more sovereign solutions to lead us, precisely, towards this transition, to accompany us towards this transition.

The fact that I, as a public official, have never heard of this tool is problematic. So something has to be done, and I intend to do something elsewhere, within my institution.

Walid: To go in the same vein, but at a higher level, that is to say at the European level, I am sharing this news. So we record on November 3rd. So on 29 October, the European Commission adopted a decision that establishes a consortium for the Common European Digital Infrastructures, the DC-EIDIC. It is a new instrument that will allow Member States to jointly deploy and operate cross-border digital infrastructures with specific governance and legal personality. It has been launched by France, Germany, the Netherlands and Italy and the DCDIC headquarters will be in Paris. This may be something to follow closely. Having seen publications, some people around the file are very enthusiastic. Of course, we’ll see, but maybe we’ll talk about it again in the podcast, who knows.

Amel: In any case, it’s true that this question of sovereignty, I think it’s critical today. And we should have more incentives. And I thought it was crazy to discover this along the way and that it wasn’t, basically, in my ENT, in fact. That it is not, by default, my workspace at the university. So, it’s good to have the choice, indeed, once again, as a researcher and as a teacher-researcher, at least at the university, we have a lot of freedom. But on this question of tools, I think that at some point, the State and then the institutions, including the University, must push for solutions that are more sovereign.

Interpersonal dynamics within a project

Walid: Okay, let’s close this chapter. The next part, we’d like your point of view as a researcher on the open source part. One of the topics we regularly deal with here on the podcast is how we manage competition, i.e. individuals within a project, so we were talking about it a little earlier, this notion of competition as well. So already, the individuals within a project… What is your view on this, on the subject? I do not know if my question is very clear.

Amel: I’ll try to take her.

Walid: Raphaël, do you want to rephrase it differently? Because there, I think I got a little lost in my question.

Raphaël: What are the dynamics of competition that you can find between people within the same project before moving on to the more external dynamic between projects? But already, interpersonal, there is this aspect. I don’t know if it’s interesting.

Walid: It’s better, it’s better.

Amel: So yes, the interpersonal aspect in open source communities. Well, I don’t think I’m telling you that. You may even be more aware of this if you contribute directly to communities. But the interpersonal aspect is central to communities. And then, the dynamics that the communities will take.

So, open source, yes, but on the other hand, in any group, group dynamics are very important. And indeed, in open source communities, it’s true that there is the component, at least for me, on the researcher side, when I approached open source, it was very: “it’s collaboration, it’s mutual aid, it’s good, let’s say”. I’m not contradicting that, but I’m going to qualify it by explaining that, well, there are also rivalries, that there’s also competition and that these are things that we don’t refuse to see, but since it’s very widespread and documented elsewhere, we don’t pay particular attention to it in open source communities, While they exist, they are there, and the competition, the interpersonal rivalries can also be very strong.

So, maybe how do you deal with them, and then why they emerge. So, they will emerge because there is also a whole mechanism of visibility, mechanisms of law also in the communities that mean that we will, let’s say automatically, in the end, put certain individuals, certain contributors in competition, in any case, those who really want to invest and evolve in the community, they will… There you go.

If we talk about meritocracy, it’s something that is very competitive, meritocracy. So, we’re going to try to give it our all, and so we’re in competition with the others, to be the one who deserves the most. So there are still mechanics that are induced by these open source communities and the way they will work. And then, there’s also the whole visibility part as an individual, then in the ecosystem, to maybe join certain companies, get hired, poach, etc. So, there is reputation that is also at stake in these communities and therefore that induces a certain dose, in any case, of rivalry that can be more or less, let’s say, tempered, managed. And one of the ways to deal with all of this is to be very clear about the rules of the game.

Amel Charleux

And then to have what we call codes of conduct or charters on what we are allowed to do, not to do, the exact processes for making a proposal, how it is studied, how it is accepted, how it is validated, how it is integrated into the software. So, try to be as clear as possible about how it works and what the rules of the game are and what the rules of conduct are. So, everything that is charter, code of conduct. It’s not necessarily always read, but it’s always good to have it to be able to refer to it. It is a kind of safeguard that allows us to be able, if there is a conflict, to have a support already written, made explicit to settle the conflict.

Raphaël: Because we were talking about interpersonal relationships, but sometimes people can be mandated or represent actors. So there are also relationships within the roadmaps or strategic orientations of the projects. Interactions between private and public actors, which do not necessarily have the same objectives. I don’t know if that’s something you observe too.

Amel: Oh yes, completely. These are things that I talk about in my thesis, in particular. This is one of the reasons why some companies are interested in paying full-time employees to contribute to projects that are open and totally collaborative. It is precisely the fact of being able to influence. So, to have influence. And to influence, you have to be better than the others, and for that, we need more developers or the best developers, who are well paid, who want to contribute to this project, who contribute well.

And so, we really see these competitive dynamics. And sometimes, what they show is competitive dynamics, not necessarily between the individuals themselves, but between the companies they represent behind them. That’s yet another level, let’s say, of competition that we can observe within these communities. But already, on the interpersonal level, what can also be interesting, we worked a little bit on this with a colleague, is the fact of having spaces for what we call doing voice. There is a researcher, Hirschman, who has worked on the mechanics, precisely when consumers are dissatisfied. So, he explains that we can either change (exit), so leave, or voice, so we need to talk, to say that we are not happy or that there is a problem. And then there is loyalty to the project, whatever the nature of the project or whatever the project does, we will always be loyal to it. And so, this space for voice is precisely to have spaces to express discontent, to express a difficulty. And that may also allow, it has to be moderate, but it can allow you to lower the tension, from time to time of voice precisely, not to keep, but to voice something, something that bothers, and to lower the rivalry a little bit.

Walid: Just, I’ll make a micro aside. If you are interested in understanding the experience of a person who works in a company that is mandated to work on a project for someone else, you can listen to the first episode on the history of GLPI where I explain my own personal case. I myself was mandated by a service company to work on behalf of the Health Insurance (CNAM) at the time, on this project, and where I explained with my colleagues at the time the arrival on the project, how it works, precisely the conflicts between the project roadmap and the internal project roadmap , etc. Anyway, I’ll close the parenthesis.

Amel: Yes, it’s an interesting parenthesis that makes me think of interviews I did and where developers told me “well, there are certain moments when I put on the project cap and then afterwards, I’m going to put on the cap of the company I work for. And I have to know when I put on which cap.” Because sometimes, it’s indeed a conflict and you have to be able to know when you’re going to push for the company or push for the project roadmap .

Walid: Indeed, the notion of reputation, I have experienced it personally. When you’re a software developer and you’re specialized in your field, finding work and you want to stay in that same job, that same software, finding work, it’s not necessarily very complicated. You just have to know which door to knock on and it can be done quite quickly…

Interactions outside the project

Walid: And so with regard to the outside of the project, there are also a lot of interactions and a lot of competition, competition between projects, between free projects, it’s a lot of work. Here, we have indeed talked about companies that can be, for example, integrators who will also compete with each other. What is your vision on this?

Amel : So, I see two different levels, let’s say, of competition and two ways of dealing with competition in your question. There is indeed competition between projects or between publishers that I would describe as of the same nature, in the sense that we know who the competitor is. So, it’s true that, in general coopetition and in competitive analyses, it’s very important to know who the competitor is.

And so, a competitor, theoretically, is someone who is dependent on the same resources and the same market. That’s what a competitor is all about. He’s someone who will use the same resources as me and who needs the same market as me. So, that’s how we manage to identify a competitor, and not just a complementer, or a potential partner, or maybe a customer or a supplier. So, you have to know who a competitor is. And I think that when we are between similar projects or between publishers, we know if we are doing the same thing, if we need the same resources, if we are on the same calls for tenders, when we see each other in the responses, do we meet in the same places… We can identify it more easily. And this competition is interesting because it is a little different in open source from what we can see in the traditional management world.

Amel Charleux

And what we were able to show in my thesis is that, in the end, what we are taught in a theoretical way — and once again, in a classical way — is that competition takes place in the market. So we compete in the market, and we cooperate on resources. We will eventually share resources, manage logistics together, perhaps. We will be able to do coopetition like this. Far from the market: the market is only competitive. And the further away we get from the market, we’ll be able to do things a little bit together, R&D for example.

Here, what we see, in the end, is a kind of inversion, a little bit, where during my interviews, I still had a lot of people tell me: “No, but there’s enough for everyone, there’s enough to eat for everyone.” So in the end, competition is not on the market, it is on resources. In the sense that, on the other hand, to have a good developer, someone who stays, who holds, who knows well, etc., that’s where it’s more complicated.

Amel Charleux

So in fact, where we’re going to have competition, ultimately in open source ecosystems, it’s a bit the opposite of what we can see in other industries. That is to say, we will compete on resources. I’m going to try to get the best of this or that technical solution, this or that developer. And then, on the market, we’re going to get together, we’re going to make a proposal together, or we’re going to get into collectives. In any case, we will defend open source together to increase the share of the pie. And finally, the competitor is rather the owner, in a way, even if there is indeed competition between open source publishers, but in the end, the big big competitor is still the owner. So it is the one we have to fight. So, come on, let’s get together on the market side. On the other hand, on the resource side, no, no, it’s me, I want to keep the developer. Or, precisely, let’s say, it’s a bit of a game between — especially for the SS2I, the ESN today, at the time — to try to have the best… There you have it, we can see how resources ultimately circulate between the different ESNs and the different publishers.

Walid : yes, it was just to say that when you’re an ESN, being able to say that you have one of the developers, one of the core developers of the software in question, it’s a guarantee of quality on the fact that the company will do things by the book, that the patches will be integrated, etc. I experienced it personally, and we won a lot of contracts at the time thanks to saying “we have core developers of software X in-house, so we know how to do things correctly”. And on this subject too, I refer people to the episode with the people at Open Source Expert. We also talk about these issues, about having the right people who are able to work and transfer the upstream code into projects. I close my parenthesis.

Raphaël : But I just wanted to add to it maybe, it can also be seen with the big open source players. Red Hat is one of them, etc. There are also like acquisitions, almost. That is to say , oh well, we hired the lead of such and such an open source project. In fact, somehow, it is a bit under the umbrella of this actor who found a strategic interest, and therefore who pays this developer or this leader, let’s say.

Amel : Absolutely, it’s a competitive advantage. That’s it, that’s clearly it. In the open source ecosystem, in any case, it is one of the axes, one of the major differentiation levers for publishers and for ESNs. And so what I was saying earlier is that I see two different things: competition between — and therefore coopetition — between publishers or groups of competing projects. And I see another level: it is the coopetition that can be established between a publisher and its integrators. And there, it’s a slightly different mechanic. We have to be a little more subtle because in the end, it’s not an identified competitor as it is. The integrator and the publisher, both are normally complementary. They are supposed to… they are partners. Nevertheless, we can observe that sometimes, we can step on each other’s toes a little, and that the publisher can offer training in the same way as one of its integrators. And so, how do we manage these relationships? How can we not step on each other’s toes? How to properly define the value axes of each? It’s really essential to manage the competition well at this level.

Walid: That’s clear…

Amel:

This is different from competition between publishers, where you are clearly a competitor, and you know who does what. But when you’re in the same ecosystem, when you’re a publisher and integrator of the same solution, well, you have to better define, let’s say, the roles. Better define the areas of value on which everyone will be able to position themselves. And this is where the business model, the economic model, is very, very important to define from the start. To be defined from the outset and negotiated as you go along, because a project evolves, economic models evolve, and a publisher may want to evolve its economic model. But it must always remain negotiated with its community of integrators, and that interests always remain aligned, in the sense that everyone knows on which value axes they will be able to position themselves.

Amel Charleux

Walid : It’s easier when the market is big, because otherwise, the publisher — since it’s generally the main access point — and potentially redispatching to integrators, the publisher has the temptation to keep the projects that interest him for himself. And in fact, he is resentful. In fact, the publisher keeps the projects that interest him. He will say: I lived this project. The first time I contacted the software in question, the publisher, he said “well, it’s too complicated for me, go see integrator X, he knows how to do it”. And the second time, he said “this project, I’m interested in it, I’ll take it for myself”. And so, the role of the integrator behind it, it is not obvious. Why will I be an integrator when the best projects are kept by the publisher? Indeed, it creates a lot of friction.

Amel : Yes, and it’s a more complex relationship than a simple competitive one. And it makes a more complex coopetition. This makes a more complex coopetition, because we are supposed to work together, we are partners, so we collaborate, we improve the product. We have to win too, we have to increase our share of the pie, so we have to work together for that. But at the same time, then, who takes what part? That is the whole question. So, there can be distributions by project size, it can be geographical distributions, it can be distributions by type of service. Indeed, in order to keep a healthy ecosystem, a healthy coopetition between the publisher and its integrators, it is indeed necessary to have very clear rules of the game.

Walid : Last case, I don’t know if it’s marginal or not, which is quite interesting, is the competition between the integrator and his community, I think, in open core models. When you have your community that comes with a feature, they submit a pull request to you because someone made the feature, and in fact, the editor, they are not interested in it at all, because it’s a feature that they want to put in their paid version. And so, he will refuse the submission explaining that “sorry, but in fact, this is a feature for us, a paid reversal. So, in fact, we’re not going to put it in the open source version.”

Amel : Yes, absolutely. There is this competition. Indeed, in the end, competition can be seen everywhere. But we also see it daily around us. So, it’s a very interesting concept, indeed. So, for this type of competition, it reminds me of the case of Odoo.

Walid: Ah…

Amel: Ah, did I touch something?

Walid : No, but indeed, it’s a case that we’ve already discussed here. For people who are interested, there is an episode on the OCA, the Odoo Community Association. I guess maybe that’s what you meant to talk about. I’ll leave you, go ahead.

Amel : Absolutely, exactly. And so, we are in this case where we have a publisher with its roadmap, with its own strategic objectives. So, he doesn’t sabotage the work of the community. He lets her live. But indeed, the perimeters on which he wants to make money, he develops them independently of what the community will develop. And what the community will develop is actually the OCA, the Odoo Community Association, which will carry it, promote it, organize it, etc. So that’s one of the ways, perhaps, to resolve this coopetition and this tension that can exist between the strategic economic objectives of an open core publisher and its community. Especially when there is so much success around a project and so many contributors. It’s true that Odoo is quite exemplary in this area.

The evolution of Open Source business models as seen by Amel

Walid : Absolutely. Precisely, you introduced a word that makes us think of the next part. You talked about evolution. We wanted to talk about the evolution of the ecosystem. I’m going to start with a very small remark about what you said earlier when talking about reputations, etc. We were talking about the fact of rights to a project. And that’s something I find interesting because it’s evolved over time. I think, for people who are old enough to have known source management systems like Subversion and others, where in fact, you had to be given a right to commit, you had to be told “you, you have the right to…” you can commit to the repository, to the repository”. And when Git appeared, in fact, there was much less of that, since in fact, in the end, anyone could fork the repository, make their own pull request, submit it, etc. It made things a lot easier. And it took away some of: “well, there are those who are deserving, they have the right to commit. And then there are those who are not deserving, they don’t have them.” Well, there you go. Anyway, it made me think of that earlier, when you were talking about rights. But here it is, the evolution of the ecosystem. Raphaël and I wanted to ask you the question: now, you’ve been working on these subjects for a while. What are the major developments you have seen around these open source business models?

Amel: On open source business models, indeed, it seems to me that there is an evolution towards a certain standardization, if I may use that word. At the beginning, there were lots of tests, lots of slightly different models, and almost as many models as projects, as many business models as projects. And then finally, along the way, and then the discussions progressed, and then the examples were shared, the good practices were shared, the models were reduced a little. I myself mapped four ideotypes in one of the articles. You can find them online, it is normally freely accessible.

Nevertheless, since the end of my thesis and this mapping, I have the impression that there is still a trend towards the standardization of models, towards what is done in the owner. There is a bit of a tendency towards closure, with a predominance of purely economic issues, purely economic considerations, to the detriment of the technical benefits of open source development and the product. I really have the impression that there is a little slippage, and that sometimes, I have the feeling that we may have missed something with open source, trying to standardize it too much, to standardize it too much, to institutionalize it too much — in the sense of making it something normal and classic, well, we have bended to the rules of the classic game precisely, of the economic valuation, of the needs of investors, to the detriment of what open source could initially be, that is to say, even beyond the philosophical or political consideration of Richard Stallman, but really a quality of development, more efficient, safer products. And so that, it seems to me that we have lost it a little.

Amel Charleux

To the extent that today, finally, an open source product is no longer a competitive advantage, it is no longer a differentiator. And there are plenty of open source companies, in the end, that simply don’t tell their customers that it’s open source, because they don’t care about it anymore. That’s it, that’s a little bit my perception of the evolution of economic models. What can open source be on the market today.

Raphaël: It’s very interesting to see these developments and to see that in the end, as it has become very business, we are playing with the rules from this point of view. Business has regained a bit of control over the habits that existed before and that were at the origin of these movements, and therefore of these communities and these technical elements.

Walid: But there is one thing that I still want to mention, and a debate on which I don’t have a strong opinion, but which I find very interesting, is when states start to develop open source themselves. And I’ll take the case of France — now there are European projects around OpenDesk and all that — but when the French government starts to develop tools itself, I’m thinking for example of Docs, which is a future competitor maybe to Notion or something like that. And in fact, he himself, basically, for his own agents, is developing something. It’s open source, there’s not necessarily a desire to make money out of it, since the goal is to equip state agents with that, and on top of that, it’s shared with other European states — so inevitably, integrators will be set up, people who will do business on it, etc. And so, we end up with a 100% open source product, not open core, which will compete with open core tools that do the same thing, some of which may even be European. I don’t know, what do you think of that? Have you followed all these debates around it? I don’t have a strong opinion on the matter. On LinkedIn, there has been a lot of heat between the pros and the antis, but I don’t have a strong opinion on the matter.

Amel: Is the question for Raphaël or me?

Walid: Oh sorry… for you Amel. I don’t know if you followed that?

Amel: This particular project you’re talking about, no. After that, what we have to keep in mind is that this is public money. And if it’s to compete with Notion or even Resana which competes with Teams, it’s the taxpayer who pays. And these are strategic choices, I think, that we are making. And so here, we are not in an economic logic of profitability. You don’t necessarily need to calculate an ROI at the end.

What we want is to provide a public service. And so, we are not necessarily on the same mechanics, indeed. The other side of the coin is that we don’t necessarily have the same budgets in front of us. And so, maybe that’s also why we’re being distanced. It’s because we don’t really put in the energy and the necessary resources, financial, to really have a significant strike force to compete with these big American players, etc. It’s because it’s a public service. It’s frugal, let’s say. It’s often frugal. So, there we go, we do it little by little. And finally, users, they’re still going to continue to use Teams or Notion, because it’s just easier. And perhaps, it’s the UX, often, it’s something that sins a little on the solutions that are proposed, not necessarily by the State, but in any case on projects that are out of economic logic, in any case. UX is the last thing, somewhere, that we may be working on. We’re going to work on the functionalities. But the problem is that behind it, users won’t necessarily be hooked.

I think these are good initiatives. For a personal opinion, I think it’s important. If we come back to the question of sovereignty, it is important. We have to come up with solutions of our own, European solutions. But on the other hand, you have to put money in the face. We must not just have nice speeches. And then afterwards, it’s a bit of a flop because it’s not followed, because there’s not necessarily the completely necessary energy that is put into it. And so, it makes a bit of false hopes every time. And even users, those who are of good will, will try to go there. And then finally, they become disillusioned and they return to solutions. I hope not to have this case with Resana and to be able to go all the way, and implement it well in my department. And that’s it, to get everyone on board with this solution and not be told “no, no, Teams is easier, it’s prettier”. It’s a bit of a risk in the end.

The impact of the cloud on business models

Walid: Absolutely. We had noted as a question: what is the impact of the arrival of the cloud on the evolution of business models? What do you think about it?

Amel:

Compared to the conclusion I gave earlier on the somewhat general evolution, I think that it has contributed a lot to erasing this technical value, let’s say, of open source. Since, in the end, everything is in the cloud, we subscribe to a service, we no longer have the machinery behind it — so I’m not talking about the developers, obviously, but I’m talking about the users, the users. And so finally, the value of open source in itself, in companies, we don’t see it anymore when we just have to connect to a space, the updates are done automatically, the thing changes by itself, and so we don’t see the code anymore. There is an invisibilization of the code, there is an invisibilization of the work, of the complexity, which means that in the end, the technical argument, the technical benefit argument of open source has tended to disappear, to fade away a little bit. And we are more on purely economic logics, in the end, of competition in relation to price, the number of features and not necessarily the quality of the technical solution.

Amel Charleux

Walid: Raphaël, next question, it’s for you.

The Impact of AI on Business Models

Raphaël: yes I have the same one. There, we were talking about the cloud. Now, we’re in the frenzy, let’s say, of artificial intelligence. So, do you think that AI will have an impact, or is already having an impact, on this ecosystem, on business models and on interactions?

Amel: Then without a doubt AI will have an impact.

It will perhaps, I think, already accelerate this commodification. The fact that code is a commodity and all the help that is put in place for developers to generate code in a more automated way thanks to these intelligences will undoubtedly push towards this code commodification. And so the issues around the quality of this code will change. It may also be a question of responsibility, since we will have to know who coded and who is responsible for this code. So there may be legal issues that will emerge with AI and the way it is integrated into development, ultimately into the daily life of developers, and into the very production of software. We will probably also have to find other levers of differentiation. Earlier, I was telling you that one of the levers of differentiation in open source is the developer, it’s the human resource, it’s the skill. If today, this skill is massively transferred to artificial intelligences, are we going to differentiate ourselves by saying to ourselves “I have the best”… There may be other levers of differentiation that will have to be considered and developed.

Amel Charleux

There are obviously issues related to data, datasets, and perhaps the question of openness will be more crucial on data more than on code today, on datasets rather than on code. So that’s something that seems to me to be quite important.

And finally, one last point that perhaps seems important to me in relation to these AI issues, especially in open source, is perhaps a shift from questions of openness, what is open, what is closed, but more: what is ethical and what is not. And so these ethical questions may be more important in our ecosystem, at least in the open source ecosystem today, even beyond whether the code is open or not. But this is perhaps what this code is for, by whom it is developed, for what purpose in the end, for what uses. And at a time when we still have developers, but also young people who are more and more interested in ecological issues, social issues too, perhaps these ethical issues will be important for publishers and for companies that evolve in this sector.

Amel Charleux

Walid: There are already things on ethical licenses on this subject.

Raphaël: What I remember, what I reason with, is the shift in focus from code to data.

On artificial intelligence, it’s black gold. The models, they are looking for more and more data, they say to themselves, well, we are going to produce synthetic data ourselves because we lack it now. We want something that is even more powerful. And so the issue you mention of ethics, responsibility, the use I make of AI, and what it allows me to do, in fact, it agglomerates for me, at least more on this notion of dataset. Indeed, maybe a shift in how we manage data, licenses and communities. Perhaps, I hope that we will see ecosystems develop — there are already some — on open datasets , where we have this transparency that is ultimately necessary to guarantee control and therefore the ethics of use, since we know what is inside. What was used to train a model? What is “engrammed” in the model in the end?

Raphaël Semeteys

Amel: The ethical issues — it’s anecdotal — but asking ChatGPT to generate a map of the world and see if it will rather qualify the Gulf of Mexico or… So we really have political and ethical issues with this data and what these intelligences are trained on. I completely agree with you Raphaël on this point. We need to know what the source is, in the end, from which the answers that are proposed to us are made.

Raphaël: I even asked myself questions in my spare time, to say: what would be the value and the point of doing this, of having some kind of copyleft licenses on the data. That is, if you train a model with this data, which has been provided by communities, by states, etc., with a specific goal. In this case, you have obligations that apply to what you are going to do with the artificial intelligence that has been trained with it.

Walid: A bit like open data with ODbL.

Raphaël: yes, that’s the kind of thing.

Walid: Well, the clock is ticking. Unfortunately, we come to the conclusion of this episode. We will certainly have a lot more to say, especially if we start talking about AI.

In conclusion, we wanted to ask you the question — but we’ve already talked about it quite a bit — so I don’t know if you’ll have any additional things to add. Do you see any trends emerging in the coming years in open source? Are there any additional things that we wouldn’t have said that you would like to mention?

Amel: I think we did indeed do a great tour. Personally, I still see this phenomenon of standardization on rather proprietary practices and on rather Anglo-Saxon economic logics and it’s something that dominates the market and what we see emerging. In any case, the way we conceive of an open source business today is very much in line with what is done elsewhere in these ecosystems.

After that, obviously, AI is a huge challenge because it will undoubtedly change things a lot and especially at a speed where we may have difficulty analyzing and then learning from what has just happened to try to do better. Because it may go very, very quickly and we won’t have time to think about it in the end to readjust properly. So I think that speed is something we have to be vigilant about. So we, the researchers, try to always be up to date and to follow the evolution of things as precisely as possible. But it’s true that sometimes, it goes so fast that it’s still difficult to describe, to explain that it’s already moved on to a next phase and we haven’t had time to digest what has just been done. So with AI, it goes very fast. We’ll see how it evolves, but it’s clearly an issue.

Conclusion

Walid: We’ll leave you a last word, if you want to get a message across before you say goodbye to the listeners of the podcast Projets Libres.

Amel: So maybe a word about research.

We’re going to go back to the themes at the beginning of the podcast. Maybe say that in the end, researchers and practitioners should work more together, collaborate more together. And it’s true that what you’re doing is already a good first step, at least in this open source ecosystem. It’s being done elsewhere, but it needs to be more encouraged and we need to have more of a reflex to call on researchers, so even on the computer science part, we have exceptional people, extraordinary researchers in the laboratories of French universities and even in management. And we are rather, as we have seen, going with the private consulting firms that the State calls upon, when there are resources, researchers who are there and who are already paid by the State, who can help to make people think, quite simply, to ask questions. Not necessarily to provide answers, but in any case to ask questions and don’t hesitate to come and get us if there are any themes that may interest you.

Amel Charleux

Walid: If the podcast listeners have questions, can they contact you? And if so, how?

Amel: Yes, completely. They can find me on LinkedIn, they can write to me on LinkedIn and then I don’t know if you can make my email available.

Walid: I will put in the transcript all the necessary information in this case.

Thank you very much, Amel, for taking the time to come and talk about your work, your research. We’re delighted, I think, Raphaël, you must be like me. It was super interesting. There were some really good things, certainly things that we’ll dig into in future episodes. Listen, glad to have you, thank you very much.

We look forward to meeting you in person at different events. And for the listeners, as usual, we suggest you run the episode if you liked it, to talk about it around you. You have the link to all our social networks in the podcast descriptions on your favorite podcast app or on the site in the footer. You will find all the information. I also take this opportunity to say that on the conferences around open source and AI… you can find on our Peertube server, the same, you will find the information in our different networks, Raphaël’s conferences on the subject where there are quite interesting things since it is his favorite subject. There you go.

Listen, see you soon for new episodes. We will certainly continue on this series in 2026. There you go, be well all of you and then follow the podcast. There are still some pretty interesting things that will happen. Amel, Raphaël, thank you very much. See you again.

Raphaël: Thank you. Thank you, Amel, for coming.

Amel: Thank you. Thank you very much. Goodbye.

To go further

Episode production

  • Remote check-in on November 3, 2025
  • Plot: Walid Nouh, Raphaël Semeteys and Amel Charleux
  • Editing: Walid Nouh
  • Transcript: Walid Nouh

This article has been automatically translated from the original language into English.

License

This podcast is released under the CC BY-SA 4.0 license or later

, , ,