I've gotten very helpful responses to the long thread I posted about splitting attention and effort between Twitter, Mastodon and other social platforms. The most helpful note that my essay was not written from the perspective of someone experiencing regular harassment on Twitter That's absolutely true. And if the gutting of Twitter's Trust and Safety team leads people to experience more abuse - as it is likely to do - it's important that people be able to leave for spaces where they can express themselves safely.

Long thread - buckle up. TL:DR; yes, you should join Mastodon. But you should stay on Twitter as well. What we need are more and different online communities, not just an exodus from a troubled platform. I opened an account on Mastodon in 2017 as part of my research on a fascinating experiment: an open source, decentralized Twitter alternative. There was a moment in 2017 when it looked like there might be a serious exodus from Twitter to something better. That moment was brief.

Paul Meosky, a doctoral student at Yale, presents fascinating research on encouraging civil conversations on Nextdoor - a deep interest of mine and @chandrn_ . The idea was to move conversations from Nextdoor into groups with an architecture designed to create civility. There were carefully stated guidelines, designed around the principles of procedural justice. Within this architected space, civil discourse and "virtuous" speech outweighed problematic speech. Reports of misuse of the platform decreased significantly.

Amazing paper from Beth Goldberg, Rick Sear and Yonatan Lupu (@yonatanlupu) on detecting and analyzing "hate clusters". They've found ~1900 communities in their analysis of VK, Telegram, 4chan, insta, gab and FB (since 2019) and Bitchute, Rumble, Twitter and YouTUbe since 2021. Based on a ton of handcoding, they've trained ML classifiers to analyze 7 different types of hateful speech and 4 problematic movements, including QAnon and Boogaloo. They're analyzing in terms of political and platform shocks.

Okay, like virtually everyone else who studies social media, I'm fascinated by @elonmusk's takeover of Twitter. Let me say upfront: I have no idea what Twitter will be like under Elon's leadership and neither do you. But that's the point. We already know little about how Twitter handles content moderation or how their algorithms work, and taking the company private makes it likely we will know less. What we do know is this: two billionaires will now control four of the major digital public sphere platforms.

Conor Friedersdorf is right to point out that fighting mis/disinfo has become an industry, and sometimes a mindset... and that the assumptions behind that mindset aren't all correct. I was at the same conference and also worried about emerging groupthink. If I were writing this piece - theatlantic.com/ideas/archive/… - my focus might have been on the limits of regulation to solve these problems. In other words, for all the enthusiasm of Senator Klobuchar and others, much of our current information disorder can't be regulated away.

I got an interesting email from a young Nigerian today alerting me to the problem of social shaming and microlending. He wanted me to promote his petition - change.org/p/central-bank… - and so I did a little research. (1/n) App-based microlending is popular in Nigeria, both because banks generally require collateral for loans and because the bureaucracy involved with lending can be overwhelming. So there's a flock of new lending sites springing up, like FairMoney. (2/n)

Like Jimmy, I remember the web before section 230. At Tripod.com, we were providing homepage hosting services to thousands of users. Like Geocities, our main competitor, our product was free to use, supported by advertising.

https://twitter.com/jimmy_wales/status/1339928608618536960

We needed millions of users to make the economics of the product work, and we needed our support costs to be fairly low. Reviewing every piece of content before it was posted would not have been possible - we simply would not have offered our homepage building tool.

I've been spending this fall at the Knight First Amendment Institute (@knightcolumbia), working through the ideas behind the forthcoming Institute for Digital Public Infrastructure that will be launching at @UMassAmherst at some point in 2021. The main outcome of that work is a series of essays I'm writing with Chand Rajendra-Nicolucci (@chandrn_) on the "logics" of different types of social networks we see in use and in development. We've published five episodes so far (knightcolumbia.org/research/mappi…)

At Columbia J-school this evening for a conversation on "Private Platforms, Public Discourse". Hosted by the Knight 1st Amendment Institute, the conversation is about law, journalism, tech and whether there are regulatory or anti-trust methods to govern social media platforms. Cindy Cohn of @EFF remembers an earlier internet where everyone having a printing press seemed liberatory, and wonders how we got to a moment where platforms are celebrating how widely they censor. We are less customers of these platforms, she says, "than hostages."

Excited for tonight's talk at @datasociety by Dr. Charlton McIlwain (@cmcilwain) on his new book, "Black Software - The Internet and Racial Justice from the Afronet to Black Lives Matter". Gonna thread some tweets here. @datasociety @cmcilwain Writing a history of black software was a scary thing to do, explains @cmcilwain - you can get lost in archives and never find your way out... which is a bad thing when you've got a book due. #databites

Reading an absolute gem of a paper about the origins of radio broadcasting and the weird role department stories had in the process, by Noah Arceneaux: jstor.org/stable/4092802… In the 1910s, some of the early experimenters with radio were department stores, which set up broadcast stations in the hopes of selling radio sets. Made sense, as the "selling ad time to companies" model didn't really emerge until 1922.