I was recently fortunate enough to attend this year's DWeb Camp near Berlin. The event was something like a conference for people interested in 'decentralisation' both in terms of governance systems and the network topologies that create governance affordances. It was also a kind of festival, with a focus on the interstitial, relational, cultural; a sprawling hangout for around five hundred hackers of all sorts, sleeping in tents and using composting toilets.

In this sense it (literally) had a bit more of a grassroots vibe than Atmosphere Conf which I also got to attend earlier this year, but there were more similarities than differences. In both cases the workshops and conversations were rarely constrained to just technical details, but deeply imbued with a sense of responsibility to society and the future.

But more so this time, the relationship between technology, the conditions of its production, and its eventual political impacts was a kind of troubled, unresolved, anxious question, with many different theories about how to make sense of it. I could roughly see two camps: on the one hand, ATProto and institutional protocol-maxis, concerned with pragmatism, global coordination, public infrastructure and sustainability; on the other hand, a more anarchic, grassroots assemblage of people concerned with peer-to-peer (P2P) networks, privacy and community resilience.

The most technical part of DWeb camp for me was actually on the train there, with Christine Lemmer-Webber, probably best known for the ActivityPub protocol underlying Mastodon and the fediverse, when she demoed a new code editor for an in-browser compiler she works on as part of her work with Spritely Institute. I'd seen the institute's use of cute hand-drawn mascots and playful names for otherwise quite heady projects - she joked that people get confused if it's a lab for software tools or an indie game studio.

Her argument was that the cute aesthetic is mission-critical: we desperately need culture and joy in our movements. Her Spritely demo showed off how easy it is to deploy peer-to-peer games, but an underlying context is the struggle to keep queer and trans culture alive and thriving in increasingly hostile political environments.

DWeb newspapers were distributed: one article by Sarah Friend reminds us that decentralising a technology and decentralising power are not the same thing, but the day after camp finished in conversation with @schare.space I reflected on how much part of me wants to conflate the two. I was a bit embarrassed of the media and software project I spend 5 years on being called Radical Directory, feeling that it implied a claim to the authority to catalog and make sense of political movements. When I changed the name of the platform I was working on to Organ, I wrote that Signal is made a more effective tool for activists than many more targeted projects, partly because of its integration into normal life.

In the 'Future of P2P' panel, participants reflected on the history of cryptography's move from controlled military weapon to infrastructural backbone, and the recent moves (like Chat Control) to try to walk this technology back. Protective infrastructure tends to have a longer life under capitalism when it also serves a crucial economic function - e.g. as TLS does for e-commerce.

Since I've moved to work on Roomy, and we slowly established that repression-resistant features like P2P transport and end-to-end encryption needed to be out of scope in the medium term, I've held on to the basic premise that generalist tools built with care and integrity can meaningfully change the playing field in favour of democratic social movements.

But I also know that as an underemployed mother, or really just as a proletarian, I'm inevitably very concerned about how I sustain working on the stuff I care about, and this incentivises me to make pragmatic concessions. At what point am I just projecting my own hopes and wishes onto a fundamentally apolitical endeavour?

Manuel Delanda already observed in the 1991 the applications of Gilles Deleuze's very decentralised 'rhizome' thinking to then-new swarm-like US military tactics in War in the Age of Intelligent Machines. Matthew Hodgson announced the return of P2P Matrix at the camp while Element, the business primarily supporting the protocol, receives ~70% of its funding from defence contracts. @dholms.at, head of protocol at Bluesky, soberly acknowledged in the P2P panel that 'if ATProto is successful, it will probably at some point be used in a genocide'. Unpacking Elon Musk and X's role in UK race pogroms, @laurenshof.online recently wrote:

"If the structure of permission is an emergent product of the amplification layer, and not something any single owner controls, then it is exactly the kind of thing that a decentralised social network would reproduce."

Decentralising tech is about giving up control over it - removing one more control lever that can cause harm, a lever that might have theoretically also done good. That lever is where politics happens, so committing to removing it means committing to the termination of some domain of political struggle, hoping that something better will bloom in its wake. Maybe the relevant political question is then, how can we make sure something better does bloom?

Carmel, who researches software malleability in Human-Computer Interaction at MIT, offered that different narratives and messaging around a product meaningfully change the affordances that users understand it to have. I wondered, who are we to present narratives and messaging? Is the company then a Lenin-style vanguard party-as-knower? What would a more Paolo Freire-inspired approach be here?

We have in any case already spoken publicly several times about how we hope Roomy will be used for community organising - I've heard a few people remark appreciatively that our online presence is overtly political. But to go further, this recent post by Belle cited Freire to suggest that devs in this space need to more explicitly turn away from patterns established by Big Tech, and to work harder to cultivate trust with the communities most harmed by those companies, so that the narratives and messaging we do put forward can form a direct loop with the emergent worlds we're trying to help bring into being. This resonated with my embodied experience at DWeb camp, and Christine's point about joy and culture.

In these kind of spaces we can cultivate relationships that are genuinely about human connection rather than money or business — sharing a sense of identity, music, art, interdependence, care, values. We need relationships like this to keep us on course when individualistic motives constantly threaten to pull us away, and like I tried to express in my talk in Vancouver, this is a prototypical germ of a new world in a particularly techy microcosm of the planetary dynamics we might hope to intervene in. The camp's tagline, what connects us is mostly underground, to me is a reminder of everything outside of money and markets that can and does change the world.

This is not to give in to naïve romanticism and neglect analysis of the material factors, but to acknowledge that building shared theories of change is an unending collective and cultural effort. It's less about being the most right than about maintaining diversity of thought, curiosity, critical reflexivity, and connection to our underlying social values. There are some parallels here with the lineage of what we might playfully call 'decentralised' communist thought:

If the problem was, as [C.L.R.] James put it in Notes on Dialectics, the supersession of the antinomy between “the party as knowing,” movement of communists, and “the proletariat as being,” communist movement, then knowing must be the knowing of revolutionary proletarians in the act of abolishing themselves.

The same text comments later on the unique position of those self-organised workers who stubbornly concern themselves, despite their limited leverage, with the intentional governance of social production:

Self-organization is always a matter of some section of the class that would be all but isn’t. Self-organization is never, then, the class itself but some fraction of the class subtracted from given relations. To organize for themselves and by themselves, the members of some projective collectivity must do so with each other but most importantly against their already existing organization by and for capital.” — The Future of Revolution, Jasper Bernes

We have our own theory working on Roomy about the potential long term impact of mission-driven slow-growth startups committed to the 'steward ownership' idea of enshrining a principled boundary between governance and capital (very similar in this sense to worker cooperatives), but as we're still looking for aligned investors, I'm keenly aware of how unproven this theory still is.

I believe in our work and am proud of what we've achieved so far both technically and socially, but ultimately I try not to hold on too hard to the idea that my political praxis can or should be limited to the scope of a day job. The Revolution Will Not Be Funded is imprinted somewhere in the back of my head, and I know the organising is a muscle that needs exercising. DWeb camp was inspiring for me in showing that there can be rich interconnections between diverse perspectives on tech's potential for social impact, and motivated me to contribute to the growth and sustainment of this critical underground world.