I devoured David Golumbia’s 400-page critique of cyberlibertarian digital discourse. Read. This. Book. In case you cannot (while I still recommend it!), I would like to highlight a few key insights I have taken away from it. Then, I will also touch on a few points on which I disagree – and it is wonderful to disagree with Golumbia because he is so specific and clear, so it is possible to pin down disagreements rather than having to manoeuvre handwavy arguments. The book is thought-provoking even if or especially if you do not agree with it.
Excessive debunking makes real discourse (almost) impossible: climate vs. digital
Golumbia compares digital policy debates to climate policy debates and draws out how painful it is to overcome fake solutions before being even able to talk about any potentially effective interventions. For the climate movement, the phases he names are 1) denial (there is no problem), 2) emission trading, 3) geoengineering (and I would add carbon capture and storage from the most recent debate). That means that even if a majority of people eventually agree that climate change poses a problem (after decades of industry-informed “science” aka lies and industry-led discourse such as “footprint”), it is still necessary to argue over industry-friendly would-be solutions. And doing so involves a very high risk of being considered radical, anti-technology and anti-progress. For example, Golumbia mentions the option of a uniform tax on emissions – which does not seem so radical, but even for that, one first has to debunk the industry’s more preferred options.
What are the equivalents in the digital space? Of course, 1) denial – it is all going just great. 2) would be something like “more education/more user empowerment”, giving people “control over their data” (to do exactly what?!). 3) would be transparency reports and “allowing researchers to get data to examine systemic risks”, while 4) would be “AI-led content moderation to better filter out racist/misogynist/… content” or obliging companies to not suspend accounts at will with no further recourse.
I often take part in discussions where I feel people have an excellent grasp of the problems, but then come up with interventions that are so clearly not suited to resolve the problem they described five minutes before.
The need for excessive debunking has tangible implications: I often take part in discussions where I feel people have an excellent grasp of the problems, but then come up with interventions that are so clearly not suited to resolve the problem they described five minutes before. When confronted with this, they respond, “well, but we need to consider options that are politically feasible right now.” I can only guess that they have too often had to go through the pain of debunking and therefore settle for one of the options that seem feasible while ignoring that their argument becomes plainly incoherent. And that renders the entire policy debate useless.
And then, Golumbia also points out how almost all of the key talking points of digital policy were anti-democratic from the start or, at least, remain open to capture from the right. To name but a few: free, open, open source, innovation, multistakeholderism, encryption, anonymisation, net neutrality, decentralisation, privacy, surveillance, free speech, censorship and internet freedom. The problem is not that these are “bad” concepts, the problem is that they are often not sufficiently defined, especially as concepts that are relevant for democracy in general or civil rights in particular. For example, “decentralisation” in cryptocurrencies is often misrepresented as a technical feature but obscures the fact that cryptocurrency holdings are far more concentrated than other forms of wealth. Or, “multistakeholderism” is used as an approach that is detached from democratic processes and enables capture by companies. I have used many of these terms in the past and will continue to do so, but I have become far more careful over time.
The digital realm is not superior
The internet is not a better place than the rest of the world, nor is it so special. Golumbia meticulously deconstructs many widely held beliefs and he does so citing from an impressive breadth of literature. He thereby shows that these are not merely his own, but shared observations which serve as threads that he weaves together into a bigger, colourful picture.
The three most important deconstructed beliefs for me were:
- The internet is exceptional, so laws should not apply to it in the same way. Golumbia shows that a general sense that the internet would be a “freer”, “more open”, “more democratic” space has justified pushback against all kinds of democratic rule-setting. And since much of the internet is run by companies, these are rules for companies, such as liability and taxation.
- Digital rights are somehow superior to and do not need to be grounded in civil rights. Golumbia very directly attacks many digital rights organisations by pointing out that they are disconnected from civil rights organisations and their underlying democratic legitimacy. Digital rights such as the right to privacy or to free speech are often framed in an absolutist way that negates the possibility of careful balancing between rights, informed by democratic processes. The fact that digital rights are defended by non-profit organisations obscures the fact that many of the organisations are systemically reliant on resources (money, career opportunities etc.) by powerful economic actors.
- Digital knowledge gives people (legitimate) power they should use for good. Golumbia highlights that having technical knowledge continues to be understood as a justification for giving those with this expertise more power, more of a say in how to shape digital technology. This obviously flies in the face of democratic values, but it is a persistent feature of the digital discourse and especially of rule-making for the digital realm. Being a technical expert confers the aura of expertise across a much broader, unrelated set of issues, including democracy, values and civil rights.
Disagreement: legal positivism and an insufficiently critical view of the state (of democracy)
There were various points on which I strongly disagree with Golumbia. Most of them can be summarised as what I see as an overly positive description of the democratic processes in the USA, on the basis of which Golumbia sometimes dismisses valid criticism. To name a few examples, Golumbia points out that both Snowden and Assange articulated libertarian ideas and do not deserve the praise they get from across the political spectrum. However, he questions whether they would actually be considered whistleblowers according to current US law and implicitly recognises US law as the only possible criterion for assessing the actions of Snowden and Assange. What is more, he argues multiple times that secret services are democratically legitimate, immunising them to criticism of e.g. potential overreach or lack of democratic oversight. This supposedly neutral position towards US intelligence services seems untenable given the CIA’s track record in violently toppling democratic governments across the world.
Golumbia also has an overly optimistic perspective of some current regulations such as copyright. He argues fervently against copyright critics – he generally assumes they want everything to be “free” and to let artists and others figure out how to buy food and pay rent. Here, a more nuanced assessment is needed to acknowledge that the current copyright system might not be the best possible copyright system, for example because a significant proportion of revenues does not go to artists, but to record labels, publishers and, in particular, academic publishers. Interestingly, copyright has become more divisive since Golumbia wrote the book in 2023 because AI companies seem to have found a great way to exploit “free” works to turn them into huge profits (or, well, huge market valuations for the time being). That leaves copyright critics often in an unholy alliance with these companies. Hence, copyright is likely to need some kind of overhaul, and this should be done based on clear arguments instead of preconceived positions.
…he argues multiple times that secret services are democratically legitimate, immunising them to criticism of e.g. potential overreach or lack of democratic oversight.
Wikipedia also attracts criticism from Golumbia for both not being sufficiently inclusive and not sufficiently valuing expertise. I have a personal perspective on this; while I no longer work at their policy team, almost 2.5 years I spent there gave me some insights (nothing secret) that cast a different light. What Golumbia highlights, in my view, is a tension that cannot be fully resolved: Managing who can edit Wikipedia articles in which way is a difficult question. People who work at Wikimedia are generally very aware of the fact that Wikipedia contributors are not diverse or representative of the wider population and there are various mechanisms in place to change this. I cannot fully judge this, but it is my impression that there are serious efforts to balance inclusion and expertise in a way that is mostly democratically organised within language communities. If there are different approaches I should be aware of, I would be very curious to find out about them.
There is so much more to say, but I will stop here for now.
Read. This. Book. I would love to talk with more people about it.