The Power of Decentralisation
A few years ago I spent months studying decentralised systems for a research paper, and the thing that stuck with me wasn't a technical detail. It was how often the conversation about privacy gets framed as a losing battle. We talk about data breaches, surveillance capitalism, and platform lock-in as if they're the weather — something to endure, not something we designed our way into. I don't buy that. Centralisation was a choice, made for convenience and speed, and it's a choice we can walk back.
Centralisation was never neutral
Every time we route a message, a photo, or a medical record through a single company's servers, we're making an ethical decision, even if it doesn't feel like one. We're saying: this one entity gets to see the data, set the rules, and decide what happens if something goes wrong. For a while, that trade felt worth it — centralised platforms gave us reliability and ease of use that fragmented, early internet services couldn't match. But convenience has a way of quietly becoming dependency, and dependency has a way of quietly becoming exposure.
The privacy failures we now treat as scandals — mass data harvesting, opaque terms of service, breaches affecting hundreds of millions of people — aren't bugs in the centralised model. They're what the model produces when you scale it. A single point of control is also a single point of failure, a single point of coercion, and a single point of profit extraction from other people's data. That's not a hypothetical; it's the architecture doing exactly what architecture does.
Decentralisation as an ethical stance, not just a technical one
What drew me to decentralised systems wasn't the appeal of blockchains or peer-to-peer protocols as novelties. It was realising that decentralisation is fundamentally about where power sits. When you distribute storage, computation, and decision-making across a network instead of concentrating them in one operator, you're not just changing infrastructure — you're changing who has to ask permission from whom.
That's the part I think gets lost in most conversations about this topic. People hear "decentralised" and think "harder to use" or "crypto." But strip away the buzzwords and what you're left with is a genuinely different answer to a question every technologist eventually has to face: should the people generating data also have a meaningful say in how it's handled? A decentralised architecture makes that question easier to answer well, because it removes the structural incentive for any single party to hoard control.
I don't think this makes decentralisation automatically good, or centralisation automatically bad. Ethics doesn't live in the topology diagram. But topology shapes incentives, and incentives shape behaviour at scale. A system where no single node can unilaterally surveil, censor, or monetise everyone else's activity is a system that starts from a better ethical footing, even before a single rule gets written.
The obstacles are real, and worth naming
None of this works if we pretend decentralisation is a free lunch. It isn't. Distributed systems are harder to coordinate, harder to secure at every node, and harder to hold accountable when something breaks — there's no single throat to choke, which cuts both ways. Performance often takes a hit. User experience often gets clunkier. And decentralisation can just as easily be captured by new elites — the people who run the most nodes, hold the most tokens, or control the reference implementation — if we're not deliberate about how we build it.
This is where I part ways with the more utopian strain of decentralisation advocacy. Removing a central authority doesn't automatically produce fairness; it just removes the easy target. The hard work is designing governance, incentives, and defaults that keep power spread out rather than letting it re-concentrate somewhere less visible. That's a design problem, not a slogan.
Why this still matters now
I wrote about this years ago because the privacy and policy pressures on centralised computing were already visible then. They haven't gone away — if anything, the stakes have gone up as more of daily life runs through a handful of platforms and AI systems trained on aggregated personal data. The argument for decentralisation isn't nostalgia for a simpler internet. It's a bet that resilience, privacy, and fairness are easier to engineer when you don't require blind trust in one operator to get them.
I still think that bet is worth making. Not everywhere, not for everything, but as a default instinct when we're deciding how to build the next system: ask who ends up holding the power, and whether that's actually who should.