Introduction
I’ve been learning about open source ethics recently, due to my contributions to formalising the governance model for CoMaps. One thing that keeps cropping up is long lists of fascists and their donors. Here’s one such list.
Now, I hate fascists as much as the next person. But this post goes on to list not just the major corporate donors (who ought to done due diligence, or have otherwise decided that fascism is not a disqualifying factor), but also “ordinary developers and small consultancies”.
I don’t think that’s a good idea. I would consider myself an ordinary developer, and I certainly don’t have time to do an extensive background check on every organisation or project that I would like to donate to.
To give an example, recently I joined a new project and was told that one of the sites I linked to from my profile had a history of poor moderation practices regarding racist language, so I moved off that service. But it took me a while - migrations are tricky and I’m busy. I’m very glad this person told me, but I’m equally happy that they didn’t write a blog post about how I’m supporting racists.
Epidemiology
At a recent hackathon I worked on a project where we modelled a problem (a satisfiability problem to be precise) using epidemiological tools, with the goal of identifying the most effective interventions to improve likelihood of a solution existing (satisfiability). That doesn’t make me an expert on the spread of infectious disease, but I can tell you that the four stages that people go through during infection in the SEIR model are:
- Susceptible (everyone starts in this state unless they are “patient zero”)
- Exposed (they have been in contact with an infected individual and caught the disease, but they can’t yet pass it on)
- Infected (they are now infecting others)
- Recovered (the infection is over, and they are now immune)
Sometimes the Dead state is also included but we didn’t need that for our problem.
One important thing to know is that the rate of transmission, $r_0$, determines what proportion of people who come into contact with an infected individual will actually enter the exposed state.
Fascism and epidemiology? Surely those are unrelated?
As I’ve been pondering the ethics of things like lists of “fashware”, I started wondering whether it is right to assume that use of fascist software is a bad thing. And I think the answer is definitely yes, but maybe not in the way you might assume.
I think it’s very easy to take the analogy of the Nazi Bar to an extreme and say that anyone who uses fascist software is enabling fascism. By this logic, I am a fascist because I use Fedora, which uses a Linux kernel, which uses Google Gemini for code review, and Google are clearly fascist (that link is a joke1). Hell, my laptop uses an Intel processor, stop using software that I wrote on it! That logic is clearly absurd (I mean, you can’t even run a desktop environment without pulling in LLM-generated code). To be clear, I know that most people are not saying this kind of thing. But there is clearly a line to be drawn here, and we need a model to help find it.
So that got me thinking, what if we modelled fascism as a non-transitive relation. In other words, if I give “value” (money) to a fascist, that’s bad, but it’s fine for me to receive their “value” (software). After some reflection, I decided that this approach doesn’t work great either, because by using their software, I am giving them some level of trust and control, both over my computer (because it runs their code) and over me (because I become dependent on using their software). Nevertheless, it is not easy to find someone who has never bought a device containing any products from AMD, Apple, Intel or Qualcomm, or who successfully avoids software from all the companies on the fashware list. So this model fails in both directions: it’s simultaneously too strict and too lax.
And this is where the epidemiology comes in. You don’t magically catch COVID-19 if you come within two metres of an infected individual. There’s a good deal of chance involved, as well as the strength of your immune system, the ventilation of the room, how long you were there, etc. And in just the same way, being exposed to fascism online does not automatically make you a fascist, but if I find out that someone uses a lot of fascist-developed software, that definitely increases the posterior belief2 that they are someone I’d rather avoid - in other words, if you are exposed more often, you have more chance of becoming infected.
I suspect you could also look at this from the opposite perspective: someone with fascist tendencies will tend to end up in the fascist spaces anyway. However, I don’t have experience in modelling social networks, so I will steer clear of that approach. It is also less fun than comparing fascism to a disease.
Conclusion
As usual, the solution is to consider each individual in a compassionate way, and accept that people may have their reasons for needing to use tools that they fundamentally disapprove of. At the same time, I think it is important to remain able to challenge harmful views, though preferably not by calling people out in public without a good reason. When it comes to supporting or criticising corporations that’s another matter: it takes a lot of work to determine whether a company is “evil” and even if you do become aware, it can feel very hard to do anything about it. I’m still not sure what to do about this except try your best and join in on collective action. Brush up on your social skills then join a cause that matters to you.