A computer is the equivalent of a bicycle for the mind.
(Steve Jobs, 1990)
The computer, used as a creative tool, is a wonderful thing, and the bicycle is an apt metaphor: It lets you go further, under your own power, than you otherwise could, and takes you places that you couldn’t have reached without it. The freedom to use, study, share, and improve the software on that computer is essential for those trips into the unknown. If you couldn’t make your own tools, you would merely be playing within a predefined area, not exploring anything new.
For this reason, Free Software arguably is what makes computing exciting. In particular, it was the key ingredient of the rise of the World Wide Web in the 1990s, and all the inventions and developments connected with it.
I’m here to argue that software freedom is still essential — but also that it isn’t enough, and hasn’t been enough for at least 20 years.
That’s because of some major developments:
- AWS EC2 was launched in 2006. Now instead of a computer that sat on your desk or in your rack, there was “compute” that you could simply rent whenever you needed it. AWS EC2 monetized computation in a new way.
- In 2007, Apple launched the iPhone, and a million smartphones followed. Technically, they are computers. In practice, they are so locked down that they don’t function as explorative, creative tools in the same way that the personal computers before them did. Instead of bicyles for the mind, they are at best described as a kickboard, or perhaps a shopping trolley.
- Around the same time began the rise of the walled gardens. Facebook, Twitter, Instagram and all the others, from WhatsApp through TikTok, replaced the open web with closed systems, in order to monetize your relationships.
Then things quieted down a bit, perhaps because everyone was busy wit blockchains. The next big change was the appearance of LLMs for public use. ChatGPT launched in 2022, starting the effort to monetize our thinking, let our skills atrophy, and then rent them back to us.
Radical monopolies
LLMs are a naked grab for political power. The investors behind the “AI” companies, with Peter Thiel as a sort of dark emperor figure at their center, are not shy about admitting that they’re trying to build a technocratic dictatorship; and LLMs are their tools.
In 1973, Ivan Illich described the concept of “radical monopolies”: a technology that is omnipresent, so that you can’t meaningfully participate in society without using it. There may well be many suppliers of that technology, so it’s not a monopoly in the strict economic sense. But its dominance means that everyone is forced to do things a certain way.
Cars are an example of such a radical monopoly. If riding a bicycle in your city is stressful and dangerous, that’s because the city was designed and built for cars. You’re not forced to own a car; but if you don’t, some things – like visiting your parents in the countryside – become difficult, expensive, or even impossible.
LLMs are just such a radical monopoly. They are being integrated into everything, and it will become increasingly hard to avoid them, especially if you have a job in tech that you’d like to keep. So not only will the money keep flowing towards Thiel and Musk and Andreesen and their ilk. In fact, we are currently in the process of rearchitecting our society to put their LLMs right at its heart, giving them unthinkable amounts of authority over what is possible and, ultimately, imagineable.
The technology is incidental. It just has to work good enough, most of the time. Even the money is just a means to an end; and that end is power.
Tools for conviviality
In the same essay from 1973, Illich also introduced the concept of “tools for conviviality”. For Illich, society-wide systems like health or education are “tools”. These tools can be designed in one of two ways: If they are conceived as “industrial” tools, they maximize productivity and drive individualism. In other words, they are pulling society apart. Conversely, “convivial” tools let people realize their individual freedom, in interdependence with others. We do things our own way, but we rely on cooperation. We build upon the things that others have built before us.
Free Software, of course, is an approach that lets us shape our IT landscape, and the internet in particular (Illich would probably just consider this entire sphere a “tool”), in a more convivial fashion. We rely on the programs others have written, and we share the ones we write ourselves. We help each other. We do things our own way; but we invite cooperation.
Beyond software freedom
The effort of reshaping society to be more convivial (or maybe just to keep the existing convivial aspects from being eradicated) is a political project. Many of the people involved may not see it that way, but that doesn’t change the nature of the work. We’re dealing with the question of who gets to make decisions about the technology we use. That is a question of who holds power; and the discourse about how power should be distributed is what we call politics.
But the site of that question is no longer the computer on our desk, and hasn’t been for two decades at least. Cloud computing, locked devices, walled gardens and monopolized knowledge all add additional layers of complexity. The fight for a convivial society needs to happen on and across all of those layers, at once.
We must recognize that a narrow conception of software freedom — the traditional “use, study, share, improve” — is not enough. It’s a necessary condition for convivial technology, but not a sufficient one. Even 40 years after the principles of software freedom were formulated, we all depend on proprietary, coercive services to live our daily lives, whether it’s taking the train, making a bank transaction, having a job or a social life.
We must also realize that software “freedom” is something of a misnomer. To avail yourself of this freedom, you need highly specialized skills and years of training. That is something that rather few people have access to. And while it’s possible in principle to teach yourself the necessary skills to make meaningful changes to software, that’s not a realistic goal for someone who may already be holding down two jobs, or who may be raising a family, or who may have a disability that takes up a lot of time and energy.
Software “freedom”, in practice, isn’t for everyone. And freedoms that not everyone can enjoy are not actually freedoms. These things are called “privileges”.
Free Software for a convivial society
This does not mean that software freedom is without value; quite the opposite. But by itself, it’s far too narrow to be a goal worth striving for. As advocates, we need to understand this.
To actually create a better society, we need to collaborate closely with many other people who have similar goals, but work with different fields and methods. Our most effective contribution is to let people experience that they can have agency over the technology they use.
“If you want to build a ship, don’t drum up the men to gather wood, divide the work, and give orders.
Instead, teach them to yearn for the vast and endless sea.”
(Antoine de Saint-Exupèry)
Helping someone to install Linux on their computer, showing them that this is a thing they can do? That’s giving them agency. Help them replace a proprietary, coercive service with a more freedom-oriented alternative. Solve a problem they’re having using Free Software. Show them that it’s possible to help yourself, and that they are capable of learning the necessary skills. That will allow them to find their own path.
In practice, this means taking the conversation about software freedom, and about creating a better society, beyond activist spaces. It means strengthening our involvement in all the other spheres of life, and finding people where they are.
If we want to be effective, we also need to learn to read power. When we move into a new setting or group, we need to make an effort to understand how decisions happen there, and how we might influence them. In the end, the magic ingredient is time and persistence. Like in many other areas, 90% of politics is showing up: being there when the decisions are prepared and taken.







