Notes from Redecentralize 2019 — Piper Haywood (original) (raw)

Been a busy few days with Redecentralize on Friday followed by MozFest over the weekend. Redecentralize was a one-day unconference at 4th Floor Studios in Whitechapel. The event was expertly organised by Ira Bolychevsky and her crack team.

It was a day of thought-provoking conversations and notebook scribbling. This is an attempt to decode the scribbles, make some follow-up plans, and to generally summarise the day from my perspective. There was a lot going on so I can’t cover it all, but I’m going to keep an eye out for other people’s notes via the Redecentralize newsletter.

\ \ \
\\ \ \ \ \ \ \ \ \ \\\ \\ \\ \\ \ \ \ \ \\ \ \ \ \ \\
\ \\\\\\ \\ \\ \\\ \\\\\\ \\ \\ \\\\ \ \\ \\\\ \\\\\\\\\\\ \ \\\\\\\\
\ \ \ \ \ \\ \ \\ \ \ \\ \\ \\ \
\ \ \ \ \ \

Kickoff and lightning talks

Lloyd Davis did an excellent job facilitating the event. He got everything going by describing the shape of the day and introducing us all to Open Space Technology with these overarching guidelines:

The activity started with lightning talks, watch them in full on YouTube (0:26–41:10). The talks in brief were:

╭--------╮------╮--╮ ╭--╮╭-╮ ╭-----------╮ ╭--╮---------╮ | ╭-|-----╮ ╭--╮ | ╮ ╯--|-|--╯ ╭----╯ | | ╭--╮ | | ---╮ | ╭--- | ╰--╯| ╰----╯ | ╭---╯ | ╰- * | | ╰-╮ -╯---╮ ╰-----╯ ----╯ * | | ╰---╯ | ╰-╯╭- ╭--|----╯ --╮----|--╮ ╰---|----╮╰--╯-|---╮ ╭--╯-----╮ | ╰-|-----╯ ---╮| -╯ ╰----- | | --- ╭--╯| ╰--*╯--╯ ╰------╯--╰-----╯ ╰-╯---╰--╯ ╰-----╯╰---

Sessions: business models, moderation, policy, privacy, identity, interoperability, and so much more

After the talks, Lloyd walked us through the session-wrangling process by inviting anyone present to come up and suggest a topic by jotting it down on a piece of paper. All of the suggested topics went up on the wall and were sorted in to one of three session slots, forming the schedule for the remainder of the day.

The range in topics was vast. Sessions included:

I was able to make it to three sessions, though I would have loved to get to more. I’ll focus on the preventing abuse session below and hope to cover the UX patterns and feminist topics in later posts.

◯◯◯◯◎◎◎◯◯◯◯◯◯◯◯◯◯◎◉●●●●◉◎◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◎◎◯◯◯◯◯◎◯◯◯◯◯◯◯◯◯ ◯◯◯◎◉◉◉◎◯◯◯◯◯◯◯◯◯◎◉●●●◉◎◎◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◎◎◎◉◉◎◎◎◎◎◎◯◯◯◯◯◯◯◯ ◯◯◯◎◉◉◎◯◯◯◯◯◎◯◎◯◯◎◎◉◉◉◉◎◎◎◯◯◯◎◎◎◎◯◯◯◯◯◯◯◯◯◯◎◉◉◉◉◉◉◎◎◉◉◎◯◯◯◯◯◯◯◯ ◯◯◯◯◯◎◎◎◎◎◎◎◉◎◯◯◯◯◯◎◎◎◎◎◯◎◯◎◎◉◉◉◎◎◯◯◯◯◯◯◯◯◯◎◎◉●●●◉◉◎◎◎◯◯◯◯◯◯◯◯◯ ◯◯◯◯◯◎◉◉◉◉◉◎◎◯◯◯◯◯◯◯◯◯◎◯◯◯◯◎◉●●●◉◎◯◯◯◯◯◯◯◯◯◯◯◎◉●◉◎◎◯◎◯◯◯◯◯◯◯◎◎◎ ◯◯◯◯◎◉●●●●●◉◎◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◎◉◉◎◯◯◯◯◯◯◯◯◯◯◯◯◯◯◎◎◎◯◯◯◯◯◯◯◎◯◎◎◎◉◉ ◯◯◯◯◯◎◉●●●◉◎◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◎◎◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◯◎◎◉◉◉●

Preventing Abuse in Decentralized Systems

There were about 16 of us in this session with Doug Belshaw from MoodleNet.

The real-world catalyst for the Preventing Abuse session was Gab, a social network that has become a cesspit of hateful speech and has been directly linked to some lethally violent events. The Verge has a good article on recent events involving Gab. In summary, the Gab creators and users see moderation as their biggest threat. To fight moderation, they’ve forked Mastodon. Since Mastodon is released under an open-source free software license, Mastodon can’t prevent Gab from using their software. Since Mastodon is decentralized, it has no control over the entire platform and can only undertake piecemeal counter-measures to fight abuse.

Mastodon has tried to take a few steps to limit Gab. They’ve released a statement on the issue, Gab is not included in the list of instances on Mastodon’s official site since they don’t abide by the Mastodon Server Covenant, individual instance admins have blocked Gab’s instance so that it is less likely to form part of the fediverse, and some developers of mobile apps such as Tusky and Toot! are attempting to block Gab in one way or another. Unfortunately, these steps may not have done much. Gab asserts that it has over a million users on Mastodon, though who knows how many are actually active.

These events amount to a crisis in both open source tech and the decentralized web. If behaviour on the decentralized web is as bad or worse than the centralized systems we already have, we’re screwing something up.

After a bit of an introductory conversation about Gab and related issues, the session broke in to three groups to discuss the problem from technical, philosophical, and design standpoints. I was part of the design group with five others.

The Internet as a physical space

How do you design a solution to a problem that has no precedent? We reached for metaphors. In particular, we compared the internet to physical places. This harks back to discussions at the Whyspace event at MozFest House last year.

So many questions came up. Are you a village, town, or city person? How do you deal with a jerk in a tiny village pub versus a Sam Smith’s in the middle of Soho? How do you deal with people that are just being slightly terrible versus people that are actually committing a crime? How can you take mental health issues and rehabilitation in to consideration when deciding how to handle people that behave poorly? How does the infrastructure of a place or the architecture of a space encourage or discourage bad behaviour?

Another consideration that was raised: No space or place is entirely neutral, and that can be a very good thing. Some US states usually vote Democrat, others Republican. In pubs you can tell bawdy jokes and curse as you please with some exceptions, in a nursery you’d better not. Society needs a multiplicity of spaces, of micro-cultures with their own practices and norms. Why do so many tech platforms pretend that a truly neutral stance is possible or even right? What about exploring the benefits of a deliberately non-neutral approach?

Comparing real-life spaces to the internet became a useful game. Usually conversations about abuse and moderation get pretty depressing bordering on nihilistic. This was both more fun and more productive.

Abuse on Secure Scuttlebutt

The discussion about the internet as a place brought us to Secure Scuttlebutt’s (SSB) “pubs” and then to a real-life instance of abuse on SSB. I’d love to find a writeup by someone personally familiar with the matter, I’ve only heard about it in conversation with people like Kieran at the Server Co-op meetups.

My current understanding of the incident is that one or more people joined SSB, followed a few existing users and pubs, and then started posting bigoted things. To counter this, users started changing the miscreants’ nicknames to something that creatively indicated their gross behaviour, blocked them, and added them to shared blocklists. The nicknames gave other users social proof of the problematic behaviour, and the block cut the misbehaving user out of that part of the network. Eventually, the bad users were blocked by so many people and pubs that they were excised from the wider network.

This prevents trolling and makes it harder for bad actors to gain a wide reach, but it doesn’t prevent them from using the platform for organisational purposes. They may still be out there somewhere in their own sort of cell. That’s more than a little scary.

It feels important to note that what happened to SSB isn’t a million miles away from what has happened to Mastodon. If the technical barrier to using SSB were just a little lower, if there were a few more bridges like Patchfox that connect current web users to SSB, SSB could be another platform option for a similar group of misanthropic users.

Next steps and further thoughts

As a follow-up, I’m going to read A Pattern Language written by Christopher Alexander, Sara Ishikawa and Murray Silverstein from the Center for Environmental Structure at UC Berkeley in 1977. This was suggested by Eileen in relation to considering villages vs cities, the microscopic vs the macroscopic. It’s available in full online, pretty cool. I’m also going to comb through Gemma’s The Internet as a Place channel on Arena.

I’m hoping that Gordon will write a retrospective on the SSB incident (*please* do!), I’ll certainly be reading that if so. I may also do some digging to see if other decentralized protocols and platforms have encountered harmful users. It seems inevitable, some must have. If so, we should be pooling these experiences to try to come up with some collective answers.

Can we create new open source software licenses that include moral clauses limiting hate speech and bigotry? The Hippocratic License, Anti-996 License, and Do No Harm License all look promising but are in varying draft states, read “Ethical Open Source: Is the world ready?” by Canadian law firm Torkin Manes for a good outline of the state of ethical open source software licensing. A related but separate consideration: how, if at all, do Creative Commons licenses help protect against the malicious use of content?

So many MozFest sessions were relevant to this topic as well. Ira called on us to break Big Tech open with stronger interoperability standards and regulations, Eileen taught how to remix UX patterns for distributed systems, and Danielle Robinson modelled the perhaps more mundane but extremely critical non-technical threats that open source projects face including licensing and governance. It is heartening to hear so many people talking about similar problems, hopefully we can bring the conversations together and amplify them.

Really wish I could have attended all of the Redecentralize sessions! I was particularly gutted to miss out on the IndieWeb how-to. I’ve tried to implement webmentions on my site but have a feeling I’m missing something… Will have to go to one of their events instead.

If anyone has additional thoughts on any of the above — especially if I’ve missed the names of any session co-facilitators — please share them with me! Email me at mail@piperhaywood.com. Thanks again to Ira, Gerben, Anna, Simon, and all of the other organisers. It was a great event.