Twitter's Birdwatch Is Civics Cosplay

Twitter's Birdwatch branding

Twitter launched Birdwatch today. Look at that magnifying glass! They're looking for the problems on their platform. Or at least they're looking like they're looking for those problems.

Author Note: This post was originally published Jan 25, 2021 on (the currently defunct) Byline.

Today Twitter unveiled Birdwatch, their new project to provide “a community-based approach to misinformation.” While this sounds like something that's meant to help people, its real purpose is to help Twitter's reputation — which is a shame because we desperately need the former.

Here's how we can tell.

Misinformation and the Mushy Middleground

“Misinformation” is a word that encompasses both incompetence and malice; it emphasizes an appearance of neutrality and therefore favors those with malintent. The Birdwatch post doesn't even mention hate speech nor marginalized groups. Does Twitter consider harm toward individuals, groups, or democracy important to a “community-based approach”?

Is Twitter even taking this seriously? Their announcement tweet features an explainer video with the example:

Whales are not real! They're robots funded by the government to watch us!!!

This winking reference to their history of technical incompetence mashed with the semi-parody bird conspiracy theory does not inspire confidence.

Centralized Architecture

While Twitter used the word “community-based” in the blog post, they are using the weakest formulation of the concept, one where Twitter lumps everyone on the platform together and appoints itself regent. Their use of “community” suggests they are aware of the desire for a more democratic moderation structure — one that embraces a diversity of experience and subculture contextualization. Birdwatch is not that.

Twitter Doesn't Need This Data

Birdwatch is not even moderation. It's data collection. These are just annotations that feed to Twitter's system where it makes final decisions. While extra context could be useful, Twitter has not demonstrated a willingness to act on the significant data it already collects. Nor has it laid out how this data would affect their opaque and regularly frustrating moderation process.

We know Twitter is not acting on its data because we've seen what it looks like in platforms that make an effort. For example, the Wikimedia Foundation developed automated tooling to surface abusive dynamics among their editors. Twitter has far more resources. Have they shown evidence of an effort like this?

An Absence of Experts

Moderation, automation in distributed systems, and community dynamics have complexity, but they're not intractable. People spend their lives studying these things, and Twitter could pay them to help build a workable system. Off the top of my head, Twitter could hire Dr. Sarah T Roberts, Dr. Safiya Umoja Noble, Renée DiResta, Dr. Timnit Gebru, or Dr. Mary L Gray.

Twitter did not do this. Instead, Twitter started with an experiment and is conducting liaison through a member of the University of Chicago's Center for Radical Innovation of Social Change (RISC).

The RISC site asks, “How do you want to change the world?” while congratulating themselves for refining carceral technologies. There's also a page dedicated to “radical thinking”: – Pay non-swing state voters to move to swing states – Use music tracking to detect and prevent mental health dips – The carbon individual retirement account (C-IRA)

Much like the Birdwatch concept, the point is to be seen working on something complex. These “solutions” do not address the underlying problems (i.e., racism and voter suppression, universal healthcare, and averting the collapse of a human-compatible biosphere).

So What Are You Offering?

I think this is a tractable problem, but Twitter is not the organization to do it. They've deeply violated our trust by allowing the proliferation of harm into the Web and civic spaces. They haven't even acknowledged their role in those harms. We should not give Twitter the benefit of the doubt here.

The good news is we don't need to rely on them. I co-founded DashKite to work on problems like these. Reader and eventually Civic are products that foster communities on the Web. And unlike Twitter, we're talking to experts to build sustainable systems.