10 Days Ago Richard Stallman Gave a Long Interview in French (linuxfr.org)

posted by Roy Schestowitz on Oct 18, 2025,
updated Oct 18, 2025

Dr. Stallman can speak Spanish and French. He is fluent at both. That makes him a good communicator, internationally, for the Free software movement. Many people in many countries speak Spanish and French. It helps when he speaks in their native language.

Earlier this month he spoke in French to a reasonably large site and community in France. We thought we'd make here an automated English translation of what he said:

On Saturday, October 4, 2025, the Free Software Foundation (FSF) celebrates its 40th anniversary! Four decades of defending the four fundamental freedoms on which free software is based.

• the freedom to run the program, for any purpose;
• the freedom to study how the program works and adapt it to one's needs;
• the freedom to redistribute copies of the program (which includes the ability to give away as well as sell copies);
• the freedom to improve the program and distribute these improvements to the public, for the benefit of the entire community.

To look back on these 40 years of fighting for free software, I had the honor of having an interview with Richard Stallman (RMS), founder of the GNU project, the FSF, and the initiator of the free software movement.

You will find the transcript of this interview in this dispatch, as well as the link to the video of it.

A heartfelt thank you to Richard for his kindness, patience, and kindness for this interview, which was a first for me. Between the stress and my stuttering, it took someone as cool as him. 😅

And a big thank you also to Aurore, the editor of this video, who cleverly managed to mask those famous stutters! 😉

The text of this interview is licensed under CC-BY-ND

Stéphane :
Hello Richard.

Richard :
Hello, it's a pleasure.

Stéphane :
It's my pleasure too. Thank you very much for allowing me to interview you on the occasion of the 40th anniversary of the Free Software Foundation. To begin, since the history of the FSF is inseparable from that of the GNU project, I'd like to go back to the origins. How was the GNU project born? Why this name, what were its philosophy, its technical foundations? And how was the creation of the FSF, two years later, a continuation of this project?

Richard :
I initially decided to develop an operating system made entirely of free software. In the 1970s, I was already using a free system, developed in the same lab as me, and my job was to improve it by modifying its code. Anyone could access this code; it was, in effect, free software. But this system eventually disappeared, along with the community that surrounded it.

For me, this was a huge loss. I then became aware of the unfair and tyrannical nature of proprietary software. I understood that it could never be just. To replace what I had lost, I wanted to create another free system, capable of supporting a similar community. I chose to model myself on the Unix organization: it was the best model to follow.

I was also looking for a recursive acronym for the name: GNU, for GNU's Not Unix. This added a touch of humor. I publicly announced the project in September 1983 and invited others to develop the various components, because a Unix-like system consists of many more or less independent elements.

Two years later, we were successful enough that it became necessary to create a foundation to manage funding, maintain copyright, and support the project. Thus, the FSF, or Free Software Foundation, was born.

This point is also important: in French, the distinction between "free" and "libre" is clear. This taught me to be more precise in English: for twenty years, I no longer use "free" to mean gratuitous, I say gratis. And for freedom, I always use freedom. So, there is no longer any ambiguity.

Stéphane :
In English, the ambiguity arose with software: previously, the context cleared up any confusion. But in computing, we find free programs... which are not free at all.

Richard :
That's true. And conversely, you can pay to acquire a copy of a free program. The two categories—free and free—are independent.

Stéphane :
Exactly. And so, the FSF not only funded the project, but also guaranteed copyright. Because it's legally impossible to write a license that 100% guarantees that a program will remain free. For example, if I release software under the GPL, those who use my software will have to release their modifications under the same license. But I could make the code proprietary at any time...

Richard :
There are two important things to distinguish: what the author can do, and what users, who receive the software under a free license, can do. Users are legally bound by the license: if it's a copyleft license, they can only create free versions.

But the author, on the other hand, is not dependent on his own license: he remains the owner of the code. He can therefore publish another version, including a proprietary one. But if he is a free software activist, he will not want to do this.

Stéphane :
Of course. And that's where a foundation like the FSF is essential: it guarantees that software under its copyright will always remain free.

Richard :
Yes. That's the FSF's mission: to protect user freedom. In other words, to prevent one user from gaining power over others. If freedom is for all, that means no one has power over anyone.

Stéphane:
So we understand the essential role of the FSF, in addition to the GNU project: to provide legal protection, particularly through copyright, and also to provide financial support. Because sometimes some companies publish software that they describe as "free" or "open source"—the two terms are often confused—but sometimes with the sole aim of profiting from volunteer work, before closing the code...

Richard:
I prefer not to use the terms "open" or "closed," because that refers to the logic of open source. However, I don't defend open source, I defend freedom.
It's true that most so-called open source programs are also free. But some are not, and it's important to make the distinction. Since our fight is about free software, it's clearer to avoid expressions that risk confusing the message and suggesting that it's simply about open source.

Stéphane :
So, I had another question about the beginning of the GNU project. Before the arrival of Linux... When I talk about Linux, like you, I only talk about the kernel, because there's a terrible confusion between the GNU/Linux system and the Linux kernel.
So, before 1991, how did GNU users use the system?

Richard :
They used it on Unix. It was the only way. We didn't have a free kernel yet, so no complete system. People installed GNU and other free components on Unix, to replace certain elements. But it was impossible to replace everything. That was the goal, but it hadn't been achieved yet.

When Torvalds released the first version of Linux, its kernel wasn't free. He had attended my conference in Finland but didn't follow my advice. He chose a license that didn't provide all the necessary freedoms. But six months later, he finally released Linux under the GNU GPL. That's when Linux became free.

For our part, we had already started a free kernel project, the Hurd. Initially, according to a friend's assessment, Linux didn't seem very interesting. But it progressed very quickly, while our design, too complex, posed many difficulties. Finally, we decided to use Linux as the kernel.

Stéphane :
Precisely, at that time, when Linux began to gain momentum, how were the relations between the FSF and the GNU project on the one hand, and Linus Torvalds on the other? Was there any talk, at any point, of officially integrating Linux as a GNU project?

Richard :
No. Linus wasn't very friendly toward us. I suspect our insistence on the name “GNU/Linux” bothered him. He didn't like that we refused to call our project “Linux,” as if it were his own. I think he had conflicting emotions. Sometimes he acknowledged the history, and sometimes he backtracked on it. It was complicated, and I can't guess his exact feelings.

Stéphane :
Another question: the Hurd kernel, which was originally called “Alix” if I'm not mistaken... Did developing it as a microkernel slow down its progress? Because you said that Linux was progressing very quickly. Was it too avant-garde an approach, which allowed Linux to take the lead?

Richard :
Yes, the Hurd's development dragged on for a very long time. And it encountered fundamental problems that were very difficult to solve. No one really knew how to solve them. That's what convinced me it was no longer worth pursuing.

As for the name “Alix,” it started out as a joke. I had a friend named Alix, who was the administrator of a Unix group. She joked that we should name a kernel after her. I decided to do it secretly, to surprise her.

Stéphane :
He must have been pleased.

Richard :
Yes, a little. But then events changed plans. The Hurd's lead developer preferred the name "Hurd." He relegated "Alix" to a single part of the code. A design change eventually removed that part. And then my girlfriend changed her name, and we broke up. But some people had already seen "Alix" appear in the code, so the rumor circulated, and she laughed it off.

Stéphane :
So, if I understand correctly, originally “Alix” referred to the entire core, then only one component, and this component was finally removed?

Richard :
Exactly. “Alix” referred to the part that handled system calls. But we eventually realized that there was no need for this specific layer: the C library could very well handle communication with the Hurd servers.

Stéphane:
You left Hurd development when Linux arrived.

Richard:
No, not immediately, a few years later.

Stéphane :
In the early 1990s, a lot of things fell into place. We saw the arrival of the Linux kernel in 1991, which, combined with GNU, finally enabled a complete system. As early as 1992, some companies began distributing commercial versions of GNU/Linux, such as Red Hat and SUSE.
How do you see their role today? Red Hat, for example, contributes enormously to free projects like GNOME, to which they are even the main contributors. But at the same time, in their discourse, they claim to be more part of the “open source” movement.

Richard :
Oh no. It's not accurate to talk about an "open source movement." The idea behind open source wasn't to become a movement.

Stéphane :
That's true.

Richard :
The free software movement is a movement to correct a social problem, an injustice. We say that proprietary programs must be replaced with free software, in order to free users from computing. Those who launched the idea of ​​"open source" rejected this ethical dimension. They did not want to acknowledge the injustice of depriving people of their freedom.

They present open source as something nicer, a more convenient way to develop or use software, if you want to. But they don't have the goal of correcting this injustice. So, for me, it's not a movement.

Stéphane :
It's more of a working method.

Richard :
Yes. In fact, Eric Raymond associated open source with a particular development method. It wasn't just him: Linus Torvalds probably initiated this approach. But once Eric Raymond described it in his writings, many people began experimenting with it, including the Hurd developers.

Eventually, this method became associated with the term “open source.” But in truth, the choice of a development method is independent of any moral philosophy.

Stéphane :
Absolutely. And so, in this context, open source officially began in 1998 with the Open Source Initiative. But Red Hat and SUSE were already distributing commercial versions of GNU/Linux as early as 1992. Before the creation of the Open Source Initiative, did these companies have the will to truly collaborate with the free software movement?

Richard :
They sometimes collaborated, yes. But they also acted contrary to our ethics. Both at the same time.

Stéphane :
So they already had contradictions at the time?

Richard :
I wouldn't say contradictions, because they never really adhered to the principles of the free software movement. From the beginning, they distributed a system that mixed a lot of free programs with, sometimes, proprietary programs. For us, this was a problem.

We couldn't recommend these distributions, or tell someone to "install Red Hat" or "install SUSE," if they contained proprietary software.

Stéphane :
Of course. But despite everything, given their significant contributions to free software projects, could they be considered allies?

Richard :
Yes, in a way. But it was difficult for us to know how to talk about it. In a give-and-take logic, we could have said, "Since they contribute so much, the natural reward would be to recommend their system." But for us, that was impossible.

We could not recommend installing anything containing proprietary software, as this would be condoning injustice. It would have placed us in a moral quandary.

Stéphane :
So, when the OSI was created in 1998, it was indeed a split. Some people didn't identify with the ethics of free software. What really caused this split? Was it copyleft?

Richard :
No, it had nothing to do with copyleft. It was a fundamental philosophical difference. For us, it's all about freedom and justice in the face of injustice. Forbidding someone from sharing copies is unjust and immoral. It's detestable! It should never be done.

When it comes to functional works—that is, works intended to be used—people deserve the freedom to collaborate with others. In the free software community, not everyone agreed with this philosophy, but that didn't stop people from contributing. People can contribute for other reasons, and those contributions still have value.

But philosophy remains important.

Stéphane :
So, the issue of copyleft alone couldn't have caused this split? Because for a long time I believed that free software without copyleft wasn't truly free.

Richard :
That was a mistake. Why so many people make it, I don't understand. On gnu.org , we have a list of licenses we've evaluated, and you can see which ones are free and which ones are not. You'll see that many non-copyleft licenses—which many people mistakenly call "open source licenses"—are also free licenses.

Copyright law is another philosophical question. There are two ways to respect users' freedom:

• either we require that modified versions remain free under the same license,
• or someone is allowed to publish a modified non-free version.

Copyright left is the method that requires that any modified version remain free in the same way. “Permissive” licenses, on the other hand, allow non-free modified versions.

Stéphane:
Some people find you too radical. But ultimately, I realize that in some ways, I was even more radical than you, since I wanted to exclude non-copyleft programs from free software.

Richard:
That's true, but I had my reasons. My goal was to be able to distribute a complete operating system that respected the fundamental freedom of users. And for that, non-copyleft licenses could suffice for certain components of this system.

As early as the 1980s, there was already useful free software released under non-copyleft licenses, because these types of licenses existed before copyleft. Since there was no need to reject them, I preferred to use them.

Stéphane:
I understand. But without copyleft, I feel like there's a risk: that a program will one day be closed...

Richard:
Either closed or restricted

Stéphane:
Restricted, let's say... Yes, I don't yet have that reflex of avoiding the use of "closed" language. What I meant is that without copyleft, we take a risk: that freedom will disappear.
I'm thinking in particular of macOS, based on FreeBSD. It's a free version of Unix, but protected by licenses without copyleft. As a result, Apple was able to take all that work and build a system that completely deprives users of their freedom. That's why I consider copyleft important.

Richard Stallman:
Yes, but it's worth noting that FreeBSD still exists, right?

Stéphane:
Absolutely, that's true.

Richard Stallman:
The words you used implied that Apple had taken over FreeBSD and made it proprietary. But that's not the case. Apple created its own proprietary version, but it didn't convert FreeBSD into a proprietary project.

Stéphane:
That's true.

Richard Stallman:
We must avoid this kind of exaggeration, because it leads to confusion.

Stéphane:
But even if they didn't remove it, FreeBSD still exists, it's true. However, Apple benefits enormously from the work that was done in a, let's say, "open" way, and has had the right to "close" it. And it's true that...

Richard Stallman:
“Open” and “closed”…

Stéphane:
Yes, you're right. I've internalized certain reflexes, like not confusing open source with free software. But in discussions with developers, the terms "open" and "closed" come up so often that I tend to repeat them. I have to be careful with this language habit.

Richard Stallman:
I want to make a philosophical point. The real problem is that Apple distributes proprietary software. The fact that it used free code from FreeBSD to develop it is secondary, a detail.
That doesn't make Apple's actions unfair. If Apple had hired a lot of programmers to write different code, without reusing the FreeBSD code, but with the same result, the injustice would have been exactly the same.

Stéphane:
That's true.

Richard Stallman:
FreeBSD researchers and developers wrote this code, and now Apple is using it without contributing anything back to the community. This is another moral issue, but one that doesn't directly fall under the free software movement.

Stéphane:
Okay, I understand.

Richard Stallman:

The harm Apple does is harm to all users of Apple products. Any proprietary program always harms its users. My mission is to make people understand this issue.

I don't want this to be confused with the question of whether the FreeBSD developers were rewarded as they deserved. That's another question that doesn't belong in the free software debate.

Stéphane:
Yes, it's a separate question, in a way. We can have an opinion on it, but it remains outside the free software movement.

Richard Stallman:
Yes. And imagine if Apple had paid the FreeBSD developers $100 million for permission to do what it did. Would that have changed anything about the moral issue?

Stéphane:
No, not at all. Actually, my question wasn't about compensation. It was mainly about the fact that the lack of copyleft allowed Apple to create a derivative version of FreeBSD—even though it's not really FreeBSD anymore, given all the modifications—without giving users the ability to check how the code works.

Richard Stallman:
Without the four essential freedoms that define free software. For me, distributing a non-free program is always unfair, because it deprives users of these four fundamental freedoms, which are necessary to have control over their computing.

Stéphane:
I also wanted to talk to you about Debian GNU/Linux, launched in 1993 by Ian Murdoch, initially with the support of the FSF. Debian has its own definition of free software, a little different from the FSF's. How did that work in the beginning? Was there any collaboration between the FSF and Debian?

Richard:
Yes. The FSF funded Debian in its early days. But soon, the project, which had more contributors, wanted to formulate a different definition of freedom, with the intention of being equivalent.

At the time, I made a mistake: I should have checked more carefully whether there might be any differences in interpretation between the GNU Project and Debian. The definition seemed equivalent to me, even though it was worded differently. I said, “That's fine.” But in reality, there were potential problems.

Later, when open source emerged, they took over Debian's definition. I don't know if they changed a few words, but they mostly changed the interpretation. From then on, it was no longer equivalent to that of free software. Today, there are programs that are considered "open source" but not free software, and vice versa.

I actually explained these differences in my essay Open Source Misses the Point .

But I should note that Debian, finally, wanted to include proprietary programs in their distribution, but put them elsewhere, to establish a very clear separation between free and proprietary components. And so, it was possible to recommend the Debian main section for installing a free system. But, after some policy changes a few years ago, this is no longer true.

Stéphane:
However, Debian, at the time, wanted to maintain a clear separation between free and non-free.

Richard:
Yes. Debian used to put proprietary programs in a separate section, and the “main” part of Debian could be recommended as a free system. But a few years ago, they changed their policy. Today, even the official Debian installer can include proprietary drivers. For this reason, we can no longer recommend Debian, not even its “main” section. And that's a shame.

Stéphane:
That's true. Yes, they recently changed their social contract, which some say gives more flexibility. But in reality, the reality is that now there may be some proprietary drivers automatically installed by the official Debian installer.

Richard:
And for that, we can no longer recommend installing Debian, or even the main section of Debian. And that's a shame.

Stéphane:
Yes, it's a shame. It's still a popular distribution, and it's true that it's a shame that it's heading in the wrong direction. And do you still make a distinction between, how could we put it, it's true that there's no open source movement, but to say the supporters of open source and Debian, is Debian still more part of an ethical logic than open source, or do you think that today...

Richard:
Yes, that's true. But Debian doesn't fully follow it like it used to. Too bad.
For more information on this topic, I recommend checking out gnu.org/distros .

Stéphane:
If we look back over the 40 years of the struggle for free software, there's been a lot of talk about actors who claimed to be close to the movement. But there was also a major adversary: ​​Microsoft. At the time, their model was based on selling proprietary software at high prices, only in binary versions. In 2001, faced with the rise of free software, Microsoft became concerned and stepped up attacks, particularly against the GPL, trying to discredit it.
At that time, you gave a lecture at New York University in response. Can you tell us about this period and its challenges?

Richard Stallman:
The big proprietary software companies didn't destroy free software, the movement, or the GNU system—thankfully. But I don't have many specific memories of that time. For example, the conference in New York… I've given so many over the years that I don't remember which one you're talking about.

Stéphane:
Oh, you don't remember? I thought that conference was a turning point. Microsoft, at the time, was notably putting forward the—false—argument of a supposed "virality" of the GPL, claiming that a GPL-covered program contaminated any software running on the same system.

Richard Stallman:
Obviously, that's false. What is true is that if you take a program distributed under the GPL and combine its code with a proprietary program to make a single piece of software that you want to redistribute, you face an incompatibility: you can't comply with both the proprietary license and the GPL. But two separate programs running on the same system are not the same thing at all.

Stéphane:
And today, what are the relations between the FSF and Microsoft?

Richard Stallman:
There's no connection. We're simply criticizing the malicious features in their proprietary software. We have a long list, with hundreds of examples, published at gnu.org/malware .

A program is malicious when it is designed to mistreat the user. This is a strong temptation in the proprietary model: the developer has power over the user and may be tempted to abuse them by adding features that strengthen their control.

Stéphane:
And on top of that, there's also the impossibility of correcting legitimate errors.

Richard Stallman:
Yes. Proprietary software has many negative effects, but I clearly distinguish between errors and malicious features.

Stéphane:
I would like to address the specific case of video games. You talked about it a little at the beginning: it's a field that combines different types of works...

Richard Stallman:
Yes. I distinguish between functional works and artistic works. Functional works—for example, software, cooking recipes, architectural plans, or sewing patterns—are made to be used. These works must be free.
Conversely, fiction or art are meant to be enjoyed, not used practically. In a video game, there are the programs that implement the rules—they must be free, since they are functional software. But there is also art, music, and narrative. These can remain proprietary.
For example, the source code for Doom was released, but not the art, not the music. This allowed the community to create other variations of the game with alternative resources.

Stéphane:
Yes, exactly. That's the example I wanted to mention: John Carmack released the engine, but not the artistic assets. And this gave rise to a multitude of variations. Hence my question: do you need a specific license for video games?

Richard Stallman:
No. The confusion comes from thinking of the video game as a whole, a package that contains everything. In reality, you have to break it down: the engine, which is a program, must be free; the art and music may not be. That way, the question becomes easy.

Stéphane:
I'd now like to address an increasingly prevalent topic: mass surveillance. This includes algorithmic cameras , facial recognition, but also messaging services like WhatsApp, owned by Meta. All of this raises big questions about privacy. Does the FSF plan to combat these technologies, or is this outside its scope?

Richard Stallman:
The FSF can't do much in this area. Its role is to promote free software. But if surveillance software is owned by the state, it must be free: the state must have the right to modify it. It would even be dangerous for it to depend on a private company to manage its own systems.
However, the surveillance problem doesn't go away because the software used is free. These are two separate issues: on the one hand, the requirement that the state use free software; on the other, the need to impose limits on what the state can do. The FSF, which remains a modest organization, does not have the means to directly combat mass surveillance.

I don't see facial recognition cameras on the streets. Nor do I see what WhatsApp does, or doesn't do, because I refuse to use it. Its client program is proprietary: I refuse to use it.

There are many injustices in these so-called services that exercise power over their users. As a matter of principle, I never use them. I resist.
There is free software that allows encrypted communication, and I use it. But never with a proprietary application, never via the server of a company I distrust.

And then there are other surveillance systems. For example, in France, the requirement to write one's name on a train ticket is unfair. We must fight to eliminate this type of tracking.

There are also systems whose objective, in itself, is admirable, but which are designed to identify each participant. For example, a system to reduce toxic emissions: it's a good thing to want to reduce them. But you have to be able to participate in this system to achieve its objective without having to identify yourself.

We must avoid forcing everyone to use a proprietary client program to identify themselves to a server and obtain, for example, the sticker to stick on their car.

The French government is interested in the use of free software in its ministries, and that's a good thing: it helps it escape the unjust power of big business. But it should also take care to protect citizens—and even visitors to France—from the danger of digital control. Because tracking people, mass surveillance, is extremely dangerous. We can see this in China: it's the ideal basis for repression.

And they also often require the use of proprietary programs. Such programs wouldn't be able to run on my computer unless they were written in JavaScript. But in that case, I block proprietary JavaScript and refuse to log in on those sites.

I've come up with a solution to the problem of low-emission zones. Each participating city should install, in well-marked locations, points of sale where the necessary plates can be purchased for cash, by providing only the vehicle's information. This would allow compliance with emission reduction regulations, but without resorting to an unfair digital system. A few such points, placed on the main access roads, would be sufficient for each city.

This would also avoid the trap of having to buy these plates before even entering France. And it's important that payment can be made in cash: it's a protection against surveillance and repression.

If you run out of cash to pay in a store, it's better to withdraw money from an ATM rather than using a card. Because if you pay in cash, the system will only know where you withdrew your money, but not what you bought with it. And for me, that's essential.

I never use my card for everyday purchases. I do have a credit card, but I only use it in exceptional cases, such as for plane tickets—since you can't travel anonymously—or for certain bills in my name, like those for my apartment. For medical care and prescriptions, I can pay by check. But other than these special cases, I always pay in cash.

Stéphane:
So it's mainly to avoid being tracked in your purchases, in fact?

Richard:
No, it's broader than that. The issue isn't simply about me personally avoiding being tracked. It's about resisting the general trend that imposes surveillance on everyone.
I do my part: I resist surveillance when it directly concerns me. But resisting surveillance that weighs on you or others—that, I can't do for them. Everyone must assume their part.

Stéphane:
Yes, the idea is to prevent this system from becoming too widespread.

Richard:
But it's already too widespread! There's far too much control, too much surveillance, too much monitoring... and therefore too much repression.

Stéphane:
And so, to finish, the last topic I would like to discuss with you is what the mass media calls artificial intelligence . Because there too, there is software behind it. I wanted to know what the FSF's position is on this subject today. And do you think there is a definition of an ethical LLM model?

Richard Stallman:
I must first distinguish my opinion from the assumptions contained in your question.

Stéphane:
Okay.

Richard Stallman:
I differentiate between what I call artificial intelligence and what I call crap generators . Programs like ChatGPT are not intelligence.
Intelligence means having the ability to know or understand something, at least in a narrow domain. But more than nothing.

ChatGPT, on the other hand, doesn't understand anything. It has no intelligence. It manipulates sentences without understanding them. It has no semantic idea of ​​the meaning of the words it produces. That's why I say it's not intelligence.

On the other hand, there are programs that really understand in a narrow domain.
For example, some can analyze an image and say if it shows cancer cells, or identify an insect: is it a wasp attacking bees? This is a real problem in some countries. These are really dangerous immigrants.
These programs, in their small field, understand as well as a human. So I call them artificial intelligence.

But LLMs, the great language models, don't understand anything. We must insist on not calling them "artificial intelligence." It's just a marketing campaign designed to sell products, and unfortunately, almost everyone accepts it. This confusion is already causing damage to society.

Other than that, if you want to use an LLM, you have to have the four essential freedoms. You have to be able to run it on your own computer, not use it on someone else's server, because in that case, they choose the program, and if the program is free, they have the right to change it, not you. And if you run it at home but you don't have the right to modify it, or to use it as you wish freely, obviously that's unfair. So I'm not saying that LLMs are essentially unfair, but normally they don't respect the freedom of users, and that's unfair.

And we must also recognize what they are not capable of doing: they do not understand, they do not know.

Stéphane:
Yes, it's true that it's marketing to call them "artificial intelligence." But many people find useful uses for them. I'm thinking, for example, of translation, which sometimes produces decent results. So if people want to use LLMs, what you recommend is to favor models under open license, is that right?

Richard Stallman:
Yes. We're currently writing how to adapt the free software criteria to apply to machine learning programs as well.

Stéphane:
So this is something that the FSF will publish in the future?

Richard Stallman:
Yes, but it's not over yet.

Stéphane:
We've reached the end of the questions I had planned. The Free Software Foundation (FSF) has existed for 40 years. I think there have been many positive advances. But today, computers are everywhere in our lives, and therefore the issue of computer freedom is more important than ever. Do you have a message to encourage people to join the free software movement?

Richard:
Yes. First, reject unfair technology: non-free apps, those that identify the user, those that track people from your computer or phone. Seek to replace every item for which free software exists, with free end-to-end encryption.
Also reject things that are supposed to be "in your home" but listen to you, reject products where commands go through the manufacturer's spying server, and resist systems that track their users. Pay cash when possible.
And if you're a programmer, you can contribute to the development of free programs, and you can also sign up as a member of the Free Software Foundation at fsf.org. Also check out gnu.org/help. One more thing: if you work at a university, invite me to a talk.

Stéphane:
Okay. And when you look back on these 40 years of fighting for free software, are you satisfied with the direction things are taking today, with the impact?

Richard:
No, of course not. Under the empire, things are getting worse and worse. I'm not satisfied. Sometimes I could have acted more effectively, but with what I knew at the beginning, I wouldn't have been able to do better. But I'm disappointed with the direction things are going.

Stéphane:
That's why it's important for everyone who is aware of the issue to contribute as much as possible: by informing those around them, by encouraging people not to use proprietary software and by helping people switch to free software.

Richard:
Yes. But the French must also unite to demand that the government's digital services respect free software. Specifically, that they stop transmitting proprietary programs to be executed on users' machines, and that they respect individuals' anonymity more. Because personal data, once collected in a database, will end up being abused, perhaps even by the government.

Stéphane:
With the arrival of Donald Trump, I know, for example, that the Mozilla Foundation explained that it was in financial difficulty because it lost funding it had from the government. Is the FSF also a financial victim?

Richard:
No, because we didn't receive anything from the state.

Stéphane:
So you have more independence than basic Mozilla.

Richard:
Yes, and we have more funding thanks to donations. That's why I urge everyone to join the FSF.

Stéphane:
Thank you very much, Richard, for this interview. So, we're celebrating the 40th anniversary of the FSF. I hope that for its 50th anniversary, free software will be used much more than proprietary software. We'll see.

Richard:
Can I make a pun?

Stéphane:
Of course.

Richard:
I love tea, but I only drink teas that go bad over time, because the others are awful.
Goodbye.

Stéphane:
Goodbye.

This coming week he'll speak in Germany, but not in German (he cannot; he'll use English). █