A few thoughts:

I think it bleeds through more than you (as also illustrated by that i dont really see any pro-ai collective communities on mastodon anywhere really, except maybe the terence tao cluster), but I think my main takeaway of the comment section on Scotts thread is that a lot of people want things to bleed through. Eugen, by comparing the pro-ai to puppy killers, and have that be the most liked post of the week, set the tone for that

I'm not sure if I agree with

@laurenshof

sentence

"It is the culture of the federation layer, formed through years of shared follows, shared boosts, shared blocks and shared arguments about what is acceptable, and it governs even though no institution maintains it."

whether is likely true that most of #fediverse users, today, has neuter or bad opinion about AI of whatever other controversial topic, nobody has the exact same "federation layer experience" as someone else.

Each one of us follow determined people, or topic, and silence or block other and that means that my federated timeline is different than yours.

If I'm seriously uninterested into something , say a bubble, or a continental popular music contest, or a sport, or whatever
I won't search for it, I won't comment it, I would eventually silence the topic.

I'm not going to tell all people worldwide not to talk about it.

I follow the creator of a particular good software because I think is relevant to hear what he thinks about software, that does not mean I agree on all his opinions about anything

Facts.

@fediversereport @Mastodon

But, the way to temper that kind of rhetoric is really the same as any other matter, through consensus. Conversations just like this. Your paper, raising concerns for the community and respectful discourse about them.

That said, people care about this stuff, and sometimes their opinions are expressed with stong feelings. That's the way people are.

To be clear, not talking about racism or abuse of others that violate instance rules.

3/

If the point is that 'hostility' is a bad way for the general concensus to be hammered out. Completely agree with that. The discourse in the Jenson thread got very heated, and it would be much better if we were better able to disagree without being disagreeable. Actually find that well reasoned and fact based arguments much more persuasive than invective or ad hominem, and think those who go down that road only hurt their case.

2/

Not really sure what this is saying. Stipulated that most people experience the Fedi as one big interconnected social network. That's what it is. And the fact that the prevailing attitude is a fair degree of hostility to AI, mirrors what most of the general population thinks about AI. The fact that an ungoverned and ungovernable network reflects the general consensus does not seem to be a problem that needs solving.

1/

Fundamentally the UI doesn't encourage it, doesn't foreground it, etc. Genuinely love that you're experimenting with doing that, and if sfba.social followed a similar lead I'd instance-hop in a second.

Yes, because "our" fediverse bleeds through to every one of the other million fediverses, Jaz. That's Laurens' point, I think.

What we need, and what I think you already identified with the Twit (shoot, Tŵt) app are very strong local presence. The local feed should be front and center, with fediverse-wide interactions coming to the forefront only when things escape the local bubble and go viral. A lot of that is UI. Mastodon's UI deliberately blurs the local/federated bubbles in order to chase that "public town square" feel.

It's hard to find the right balance, and as someone who like to think tech can provide the answers, I fear the answer might be that the balance is different for every single community; no one UX can correctly be applied.

cc @fediversereport@mastodon.social