acesAug 7, 2023
Evan Prodromou

Hi, fediverse Web and mobile client developer.

What if you added a speed bump in the reply flow to slow down users for just a moment?

"Do you really want to reply to X? Will they appreciate your reply?"

Maybe just the first time they reply to someone; maybe easily turned off in settings.

If it makes people think for just a moment, it might help with making the fediverse a more welcoming place for women.

UPDATE: deleted and redrafted per multiple requests.

Aug 6, 2023 at 4:20pmWeb
Show threadEvan ProdromouAug 7, 2023

This may help with some angry or harassing replies.

However, I'm more interested in the high volume of microaggressions that women and nonbinary people report on social networks, such as:

* Centering yourself
* Hijacking a thread to change the topic
* Being overly familiar with someone you have a parasocial relationship with
* Multiple replies "thinking out loud"
* Talking down to someone who knows as much or more than you do about the topic

Would a speed bump help here? Maybe.

Show threadEvan ProdromouAug 7, 2023
I think an opt-in mode, off by default, to help people remember to better regulate their own behaviour, would have benefits without inhibiting marginalised people who have issues speaking up and sharing their voice.
Show threadEvan ProdromouAug 7, 2023
Also: *an* answer is not *the* answer. There are other tools that need to be here, including reply hiding (take this reply out of the collection of replies to my post) and reply control (only friends can reply, only followers can reply).
Show threadCharles RoperAug 8, 2023
@evan Mmm, I like that. Thread curation tools. Reply pinning could be good (pin the good ones to the top) and reply sorting (by likes). Kinda like StackExchange.
Show threadJessamynAug 7, 2023
@evan I think "speed bump" is exactly the right way to look at it. Maybe something like after 2 or 3 replies, you have to wait longer between replies unless both parties opt in to continued conversation? Like. masto is great for long back and forth topics, but it's also a little tense when you get that "reply guy" type of interaction, people who won't quit. We've talked about and never implemented this for MetaFilter but I think about it a lot.
Show threadSean BalaAug 7, 2023

@jessamyn @evan

Really love the idea of an enforced gap between replies after you pass a certain number of replies.

Show threadAnnie Hsh ðŸ––Aug 7, 2023

@evan Yeah, I think an optional tool that people can choose to use if it makes them feel safer to post couldn't hurt. Even the fact that people know that they have some additional control over the replies they receive could be beneficial.

Objectively, I don't think any of us believe it's going to solve the entire issue, that's why we are having these discussions and point out why these behaviours are undesirable in the first place. But if a small improvement is achieved, it's still worthwhile.

Show threadLyle, a walking haunted houseAug 7, 2023
@evan I might be overly cynical about this, but "opt-in, off by default" things like this seem unlikely to make a huge change. It prevents people from accidentally being an asshole when they don't mean to, but would do nothing to curtail people intentionally being an asshole.
Show threadEvan ProdromouAug 7, 2023

@x1101 Agnes shared some good results.

https://pdx.masto.host/@agnes/110843878008954696

Very high response rate! It could be a very successful intervention.

Agnes (@[email protected])

@[email protected] I’m certain you’re already aware, but when #Twitter did the experiment, this was their finding: “if prompted, 34 percent of people revised their initial reply or did not reply at all. After being prompted once, people composed on average 11 percent fewer offensive replies in the future“ The fediverse can’t be dominated by a single app, so I think unfortunately all the mean people will self-sort to use an app that will allow them to be mean as easily as possible 🙃 https://www.nbcnews.com/tech/social-media/twitter-begins-show-prompts-people-send-mean-replies-rcna839

Show threadLyle, a walking haunted houseAug 8, 2023

@evan That's a better improvement rate than I'd expected, but it seems like she drew the same conclusions that I did?

Intentionally bad actors will ... act badly.

I'm not suggesting that that's a reason not to try to make our communities a better place. I just don't see an "opt in" way for that to happen.

I don't see a good answer, but I am glad that we keep coming back and thinking of / trying new things.

Show threadSamir Al-BattranAug 6, 2023

@evan
But Evan, do you think the trolls and a'holes will care? I think they would simply say yes and troll as usual
While creating a friction for those who actually care and wouldn't say anything mean

So IMHO this is counterproductive

Show threadMans RAug 6, 2023
@evan There's an xkcd for that. https://xkcd.com/481/
Listen to Yourself

xkcd
Show threadTim ChambersAug 6, 2023
@evan There are a ton of #calmtech projects that should find a home in the Fediverse as the ever present push for more clicks, more engagement and more activity that commercial social needs to sell ads simply doesn’t apply here.
Show threadEvan ProdromouAug 6, 2023
@tchambers what a good point!
Show threadtechnicatAug 6, 2023
@evan I have a reply confirmation step in my app.
Show threadEvan ProdromouAug 6, 2023
@technicat any negative feedback so far?
Show threadtechnicatAug 6, 2023
@evan I haven't released the app yet, so far just figuring out what I want in it. I have a confirmation for posting in general - I've found, like you said, if I take a pause before sending something out, I may notice spelling or wording or tone that I want to change, or I might just change my mind about sending it.
Show threadthepixelfoxAug 6, 2023
@evan Honestly, the whole thing about making it more welcoming for women, I get it. People online can suck. But As a woman, when reading that it feels like we're still viewed as some fragile, breakable objects who can't think for ourselves so we need others to implement was for us to be safe. There's block and report buttons, and we have the ability to stand up for ourselves.

And trying to, politely, dissuade someone from replying to another person feels like a very mild form of control, trying to change their thought pattern so they won't do something. It's a mind game and I'm not sure how to feel about that.

Yes, there's some assholes who go around being jerks to people, you'll get that everywhere. But you can just block them. Rather than having a programme that tries to alter a thought pattern of someone.
Show threadSamir Al-BattranAug 6, 2023
@thepixelfox @evan
unfortunately the lack of algorithmic feeds (and notifications) makes blocking not as useful.
For every person that blocks them, trolls will find thousands more to annoy.
Though I agree with you that they will simply use an a client that doesn't ask them. I think moderation on good instances, and fediblock for bad ones are more effective in those cases
Show threadEvan ProdromouAug 7, 2023

@thepixelfox that's a big part of what social network software does; encouraging or discouraging various social interactions.

Every affordance we have in our apps encourages using that feature; every constraint discourages it.

If we'd like people to slow down on their replies, a constraint may be helpful.

I don't think the needs of the individual are paramount in a social network. We're all here, and we all have to deal with the behaviour. Nudges in the right direction help everyone.

Show threadOcéaneAug 6, 2023
@evan Here's an alternative idea: develop software built around communities first, and features second ; or around the UX first, and features second ; instead of building software around giving dopamine to people not getting it from their productivity, following a libertarian ideology of excuse, because they would lack the economic/cultural/social/capital to develop personal projects, if not a career around them.

Don't develop software that's actually ableist and anti-poor (and thus, racist and misogynistic). This is how you can actually help women.
Show threadmlinksvaAug 6, 2023
@evan I'm for more friction. Prefer sender pays, but would appreciate this!
Show threadEvan ProdromouAug 7, 2023
@mlinksva it could be exponential like a Donald Knuth bug bounty; your first reply is $1, 20th reply is $1M+.