Why the problem with social media isn't the algorithm — it's the ownership structure
Over lunch on the final day of Rebuild 1, we sat down with Niels Oerbaek Chemnitz and Mikkeline Sofie Skjerning Thomsen from Os & Data, a research cooperative that explores digital and social media in Denmark, to discuss theories of platform ownership outside the accepted, standard models.
Os & Data's interest in social media is focused on the ways in which different platform structures appear to incentivise different types of behaviour. They have studied Facebook usage in Denmark, in particular, and found that some of the different structures within Facebook create very different types of conversations: Pages tend to lead to negative interactions and outcomes, while Groups tend towards positive, constructive and supportive dialogues.
Inspired by these insights, Mikkeline Thomsen says that their goal is to 'initiate a project to create a democratically owned alternative to the Facebook Page Group. We're a democratic collective — everyone is an equal owner, none of us can make an exit or pull out profits, and so we have "tied ourselves to the mast", so to speak; we would like to see more builders taking the same approach to the social platforms they are creating.'
As she puts it, 'it's the elephant in the room. We are very keen to discuss and criticise the dire state of the platforms and the effects that they have on people, but we seem less willing to draw a throughline from that to the ownership structures that create these conditions and lead to the platforms being the way that they are. We can't just build the same systems and structures as we have had for the past 15 years and which we all agree are making the world sick, but say "no, this time they are good because they are European!"'
Niels Chemnitz shares this scepticism as to how much will change without altering something fundamental about how companies are structured. 'How do you avoid making the same mistakes in the future? Do we want to be here in 20 years' time, talking about the European tech founders of the 2020s, and how it turned out that they were not in fact the messiahs? That they ended up succumbing to market dynamics and investor pressures to pursue growth at all costs? Just like the founders of the 2000s and 2010s did?' As the old saying goes, 'madness is doing the same thing over and over again and expecting different results'.
They both believe that what is needed is a fresh approach not just to platform functionality but to everything: ownership, revenue models, the underlying technology, and the algorithmic layer.
'From what we've seen,' he adds, 'the most valuable product for democracy is not a new or interesting product; the interesting thing would be a different governance and ownership structure which removes, or at least limits, some of the incentives for the exploitation of users.'
Perhaps counterintuitively, there are clues in Facebook Groups themselves. ‘Self-governed communities like Facebook Groups are the old promise of the internet. The most popular infrastructure is basically still self-governed, self-organised, many-to-many communities — forums and groups, collective spaces. These are popular because they have not been iterated on, have not added exploitative or extractive layers, and there are lessons to be learned from the way these are structured.'
So how might alternative ownership models work? Mikkeline Thomsen floats several options: 'There are lots of alternatives, like industry foundations, where the foundation has its "bible" which defines its purpose and its ethos; in this example, power is decentralised, which means that no one party can roll back the purpose unilaterally. This guards against abuse and guarantees certain standards of practice. Or you could create an association ownership structure, where the platform is collectively owned: we all own it and no one does, it's owned by the association as a whole meaning no individual has executive control. Another option is a mixed ownership structure, whereby there is a shareholder structure in which the majority shareholder is an association, which, again, prevents a single actor from taking the company in the wrong direction.'
'One of the consistent themes at Rebuild has been that social platforms should by now be considered critical infrastructure; maybe, then, they shouldn't be governed by capitalistic systems. Could such infrastructure be better served through a democratic ownership structure? We have that a lot here in Denmark, for example, I am a "member and owner" of my insurance company, I am engaged in electing the board… structures like these both create investment in the ethos of the business from users and guard against executive malfeasance.'
They are both keen to stress that this doesn't have to be anti-profit. 'This doesn't have to be pro bono or charitable — people can make money. But what we see is that we often have profit-sharing systems, so that there's no extractive revenue model or revenue maximisation. There's no growth for the sake of growth if that means making a worse product, and no one is demanding a hockey-stick adoption curve or a 10x exit.'
It's also possible for these alternative models to attract other sources of funding. 'Because we want to make a not-for-profit platform, we can attract philanthropic foundation funding, which feels like a positive alternative funding model; we can make the case that this is supporting civil society through infrastructure. One thing that VCs can't buy you is users — you can acquire a platform, or buy ads, but that's it. If we offer the users ownership, that's a potential incentive for them to transfer over. We are trying to do this with moderators of Facebook Groups in Denmark right now — they will get a share of what we are building, so they will benefit from it through things like a say in how the platforms are moderated, for instance, or through reduced costs.'
Niels Chemnitz was keen to stress that these models don't need to be publicly funded. 'We don't always have to have state funding… we can be private, just not extractive. There is a space in between that people can explore. This is something that fits with Danish society. For example, here we have Danish Crown, a huge company but one that's democratically owned by the farmers, despite its profit model.'
'This is about moving to a post-billionaire model; historically, starting new platforms required lots of capital and needed investment, but that era might be over; the development costs are falling thanks to AI, building is easier and faster and talent is less scarce. There's also a lot to be done with the tools we already have. We don't need lots of expense, it's more user-centric and less capital-intensive.'
Both Niels Chemnitz and Mikkeline Thomsen see these alternative structures as potential ways to incentivise the 'shift' from US platforms to other alternatives; co-ownership and influence on digital spaces are useful carrots to attract users to make the leap. 'What is our motivation? It's to fix the fact that these platforms feel like they are making society sick, and the systems that have brought us here haven't been working — so we need better systems. We all seem to agree that there's no need for a single, "big European platform", and that we need an ecology with interoperability, which in itself is a pluralistic and democratic approach — so why not try collectivism all the way down?'