from the power-to-the-people dept
We just wrote about Substack’s issue with content moderation and the Nazi bar problem. As I highlighted in that piece, any centralized service is going to be defined by their moderation choices. If you cater to terrible, abusive people, you become “the site that caters to terrible abusive people.” That’s not a comment on “free speech” because it has nothing to do with free speech. It has to do with how you keep your own corner of the internet and what people will associate with you and your brand.
This is why I’ve argued for years that any one particular private service cannot be “the town square.” The internet itself is the town square built on decentralized protocols that allow everyone to communicate. Each centralized service is a private plot of land on that wider open vista, and if you want it to be unkempt and full of terrible people, you are free to do so, but don’t be surprised, or act offended, when lots of people decide they don’t want to associate with you.
This is also why I’ve spent so many years talking up the importance of a protocols not platforms approach to free speech. With decentralized protocols, the questions are different, and the ability to speak freely is retained, but the central issue of abusers, harassers, nonsense peddlers and more can be dealt with in different ways, rather than looking to a centralized nexus of control to handle it.
This is why I remain encouraged about Bluesky, the decentralized social media protocol which was ever so slightly influenced by my paper. It’s been in beta testing over the past few months, and has plenty of promise, including in overcoming some of the limitations of the ActivityPub-driven fediverse.
Around the same time that Substack’s Chris Best was melting down in response to fairly straightforward content moderation questions, Bluesky put up a blog post explaining its philosophy around content moderation: Composable Moderation.
Moderation is a necessary feature of social spaces. It’s how bad behavior gets constrained, norms get set, and disputes get resolved. We’ve kept the Bluesky app invite-only and are finishing moderation before the last pieces of open federation because we wanted to prioritize user safety from the start.
Just like our approach to algorithmic choice, our approach to moderation allows for an ecosystem of third-party providers. Moderation should be a composable, customizable piece that can be layered into your experience. For custom feeds, there is a basic default (only who you follow), and then many possibilities for custom algorithms. For moderation as well, there should be a basic default, and then many custom filters available on top.
The basics of our approach to moderation are well-established practices. We do automated labeling, like centralized social sites, and make service-level admin decisions, like many federated networks. But the piece we’re most excited about is the open, composable labeling system we’re building that both developers and users can contribute to. Under the hood, centralized social sites use labeling to implement moderation — we think this piece can be unbundled, opened up to third-party innovation, and configured with user agency in mind. Anyone should be able to create or subscribe to moderation labels that third parties create.
The actual details of how this will be implemented matter, but this seems like the right approach. There is certain content that needs to get taken down: generally child sex abuse material, outright commercial spam, and copyright infringement. But, beyond that, there are many different directions one can go, and allowing third parties to join in the process, opens up some really interesting vectors of competition to explore alternative forms of moderation, and create different views of content.
Here’s the way we’re designing an open, composable labeling system for moderation:
- Anyone can define and apply “labels” to content or accounts (i.e. “spam”, “nsfw”). This is a separate service, so they do not have to run a PDS (personal data server) or a client app in order to do so.
- Labels can be automatically generated (by third-party services, or by custom algorithms) or manually generated (by admins, or by users themselves)
- Any service or person in the network can choose how these labels get used to determine the final user experience.
So how will we be applying this on the Bluesky app? Automated filtering is a commoditized service by now, so we will be taking advantage of this to apply a first pass to remove illegal content and label objectionable material. Then we will apply server-level filters as admins of bsky.social, with a default setting and custom controls to let you hide, warn, or show content. On top of that, we will let users subscribe to additional sets of moderation labels that can filter out more content or accounts.
Let’s dig into the layers here. Centralized social platforms delegate all moderation to a central set of admins whose policies are set by one company. This is a bit like resolving all disputes at the level of the Supreme Court. Federated networks delegate moderation decisions to server admins. This is more like resolving disputes at a state government level, which is better because you can move to a new state if you don’t like your state’s decisions — but moving is usually difficult and expensive in other networks. We’ve improved on this situation by making it easier to switch servers, and by separating moderation out into structurally independent services.
We’re calling the location-independent moderation infrastructure “community labeling” because you can opt-in to an online community’s moderation system that’s not necessarily tied to the server you’re on.
This, combined with Bluesky’s plan to allow anyone to create their own algorithms, and offer up a kind of marketplace of algorithms, is what makes Bluesky such an interesting project to me, in that it creates a much more decentralized social media, but without the philosophical issues that often seem to hold back Mastodon (some top down decisions and norms against any algorithms or search, and still relying on individual instances to handle moderation issues).
I’ve seen some people complain about the initial implementation of Bluesky’s content moderation system, which is in user settings, and pops up a window like this (with these defaults):
The negative feedback I heard was that setting things up this way suggests that Bluesky is “okay with Political Hate-Groups” but I actually think it’s much more interesting, and much more nuanced than that. Again, remember the idea here is that it’s allowing lots of people to put in place their own moderation rules and systems, allowing for their to be competition over them.
This approach, actually has some pretty clear advantages, in that it gets us somewhat past the nonsense about “censorship” and basically says “look, we’re not taking down your speech, but it’s not going to be seen as the default.” And, on top of that, it takes away the excuses from huffy nonsense consumers who whine about not being able to get content from their favorite nonsense peddlers. You can argue that nonsense peddlers should never be able to find a space to spew their nonsense, but that’s never going to work. Beyond being an affront to general principles of free speech, it also is simply impossible to stop.
We’ve seen that already: people banned from Twitter or Facebook found their own places to speak and spew their nonsense. That’s always going to happen. With a setup like this, it actually does help limit the biggest concern: the discoverability that drives more users down the path from reality to nonsense.
But, also, importantly, a system like this actually makes it easier for those who need to monitor the nonsense peddlers, from law enforcement to academic researchers to the media, great visibility into what’s happening, and to have better responses prepared.
Again, these are early days, but I’m encouraged by this approach, and think it’s going to be much more interesting than lots of other approaches out there for content moderation.
Filed Under: at protocol, composable moderation, content moderation, decentralized, protocols
Companies: bluesky
Source link
