@3wordchant I think it is a good point that this may require a policy change.

Since Gab does not have a recognizable hate speech policy, no, I don't think a large number of users there would change our approach.

I think the availability of a tool like limiting allows us to make an appropriate choice. In my view social media should never be entirely on-or-off, just like social life should not be. People are complex, and the fediverse should reflect that.

@ntnsndr there's evidence to suggest that sanctions are effective at reducing harmful behaviour¹; it doesn't simply migrate elsewhere when communities are banned in one place. For me (and by my reading of s.c's Federation Abuse Policy), that applies equally to "Facebook should ban hate groups" and "s.c should not federate with Facebook until they enforce their rules against harmful behaviour" (1/2)

¹techcrunch.com/2017/09/11/stud…

CC @flancian

@ntnsndr more evidence about the efficacy of deplatforming far-right groups: idz-jena.de/fileadmin/user_upl…

CC @flancian

@ntnsndr thanks for the paper, I'll read it as soon as I can.

Meanwhile, I fully agree with you that it will be helpful to have better tools.

In the mean-time, while those tools are built, and given that you are familiar with the evidence of deplatforming in reducing harmful behaviour, why do you think it is better not to use the (sure, blunt) tools available? (1/2)

CC @flancian

@ntnsndr Do you think there might be some component of personal privilege in your preference for this approach, and do you see any potential exclusionary impact in applying that preference to a community space that intends/claims to be welcoming to people who face more marginalisation than you do? (2/2)

CC @flancian

@jotaemei @3wordchant I think our rules and moderation team are wise enough to detect the difference between affirmation and critique.

Thanks for these points.

@jotaemei off top (others with more knowledge of ActivityPub might be able to think of further examples), defederating would prevent Threads users from organising harassment campaigns invisibly in replies to s.c users, and prevent s.c. users' content from reaching unexpected audiences of hate groups on Threads by being boosted (or whatever Threads calls it there).

CC @ntnsndr @flancian

@3wordchant @jotaemei @ntnsndr these are good examples, thank you. Playing devil's advocate here a bit:

- The telephone can be used to organize a harassment campaign. Should numbers not be able to call each other freely because of this? Should the government tap all lines because of this? My gut feel says no to both. Does this intuition not apply here because of speed or some other factor in this particular network? I'm unsure.

@3wordchant @jotaemei @ntnsndr

- On boosts as a danger/weapon: I'm sorry but I don't see how federation makes the problem significantly worse for what amount to public web posts that can already be scraped, etc. Maybe a visibility rule to 'only show to logged in users from instances in a user-kept allowlist' would be needed for such cases?

Essentially user-defined per-post federation allowlists might be needed in the long term.

"authorized fetch" is part of what you're describing, and I hope its adoption continues to increase.

As for "you can still see the content on the web", sure, but there's a wide zone between "technically impossible" and "absolutely trivial to do" – surely you agree that putting *any* friction in the way of the bigots who demonstrably exist on Threads will reduce the amount of harm caused, even if not to zero?

CC @jotaemei @ntnsndr

@3wordchant @jotaemei @ntnsndr My (surely privileged, tech-bro-influenced position) is currently "open web first", and if someone doesn't want their posts to be seen widely they should use a non-open visibility setting.

This doesn't mean I think we shouldn't defederate with actively fascist instances, or we shouldn't work to improve the paltry visibility settings we have now in Mastodon. We should do both. It's just threads doesn't seem like a fascist/troll instance to me, and I've seen plenty.

@3wordchant @jotaemei @ntnsndr now, if once threads has set up moderation/admin communication channels harmful accounts stay up... then my position about them will change.

You pointed out earlier that this position might be inconsistent/irrational as the onus of work should be on them given their track record. That's fair. I'm still processing this and I might change my default position because of this.

Exactly. In February, Facebook will celebrate 20 years of having had the opportunity to set up effective moderation. The parent company's 2022 revenue was over $116 billion; Instagram (the business unit of which Threads is a part) had estimated revenue over $50 billion the same year. I think it's very fair to say that they have had a huge opportunity to improve their content standards, if they were going to.

CC @jotaemei @ntnsndr

@ntnsndr while I agree with you about the benefits of decentralisation, I think framing this as a "basic fact" about scale ignores the factors specific to FB's organisational structure, constituency of its investors, business model, and the nature of the (lack of) legal regulation in its country of origin, and many of the countries where it is most popular.

CC @flancian @jotaemei