De-escalating social media conflict • Hacker News Board

Re: De-escalating social media conflict

#121

Post by earthboundkid » Sun, Jul 05, 2020, 1:54 AM UTC

ssalazar wrote: post #86

noobermin wrote: post #57

I think this work is great and interesting, and I respect it. One of the things I feel like it lacks is a discussion of the broader context: we all sort of suspect that one of the reasons Twitter does not generally allow for de-escalation (like FB not censoring misleading news articles) is that such escalation is in fact part of their revenue stream (escalation=more engagement), so while I think ideas like this would…

"Engagement" is the "paper clip maximizer"[1] of the modern software industry, or more generally an example of "instrumental convergence." Optimizing engagement without regard for negative externalities is a catastrophic ethical failure of the major players in the software industry, and we are finally seeing a backlash both from mainstream advertisers and widespread public outcry. [1] https://en.wikipedia.org/wiki/In…

Capitalism is the paper clip maximizer, non-metaphorically.

Science fiction is pretty much always just about things that already happened. “What if we maximized a non-human value?” is just a subconsciously repressed way of asking “Did anyone notice we’ve been maximizing non-human values for a couple hundred years?”

Re: De-escalating social media conflict

#122

Post by joe_the_user » Sun, Jul 05, 2020, 1:54 AM UTC

npunt wrote: post #98

joe_the_user wrote: post #66

Earlier quoted context omitted.

Well, if you have someone determined to go after someone else, the most nuanced approach is not going to get you much. The attacker just rephrases it as they wish. Which is to say, when you have a medium like Twitter, where the entire world can march in and get involved with conversation X, a proposal for de-escalation seems futile and a bit absurd. Among other things, a lot of media personalities have built their en…

Twitter isn't just one thing, it's composed of many communities of various sizes and closeness. Some communities are toxic and enjoy kicking people when they're down, others less so. We tend to recall only the extremes of these big personalities doing bad things. People model behavior every day on Twitter. If people only see or recall others kicking people when they're down, that's what people will expect is the norm…

Twitter isn't just one thing, it's composed of many communities of various sizes and closeness.

That's true tautologically in the sense that different people follow different things and interests tend to cluster. But Twitter doesn't have communities that can easily kick people out of their entire group, IE, the only enforcement comes from Twitter's own lax rules.

People model behavior every day on Twitter.

Sure, on average. But determined person, say a troll, is in no way constrained to model anyone's behavior. That's kind of general problem of bullies - most people model the average and bully does what they want, so the average moves towards the bully over time.

Re: De-escalating social media conflict

#123

Post by oehpr » Sun, Jul 05, 2020, 1:55 AM UTC

kortilla wrote: post #99

jacquesm wrote: post #56

Earlier quoted context omitted.

Exceptions, rules and so on.

Ironic response :)

Well, give jac their fair dues. They said "You can see that dynamic at work here on HN on a daily basis." not "Every single post where someone admits a mistake on Hacker News is guaranteed to be capitalized on by outrage seekers"

Re: De-escalating social media conflict

#124

Post by npunt » Sun, Jul 05, 2020, 2:01 AM UTC

memexy wrote: post #113

> Reply to your tweet, posting "I was wrong" There is another option. Quote tweet and say "I was wrong". Quote tweeting is a concept extension mechanism but people mostly use it for dunking. If you make a mistake then quote tweeting is the best way to correct it because people can then search for "I was wrong" and see the wrong tweet in context.

True, quote tweeting (I think you mean retweet with comment) yourself is another option, however I don't see it as fundamentally different to replies.

If people are liking, retweeting, or replying to the original tweet, it will continue to show up in the feeds of others through the home feed algorithm. Thus, the same forces apply that strongly favor the original over the correction.

Re: De-escalating social media conflict

#125

Post by oehpr » Sun, Jul 05, 2020, 2:03 AM UTC

Fellshard wrote: post #2

How will this play when a mea culpa acts as blood in the water for some groups of bad actors?

Currently I am liking the look of this proposal a lot! I've been trying to imagine some scenarios where it could be used or abused in a negative way but so far I haven't come up with any.

Could you go into details on how you're thinking this would play out? I'm not sure how a bad actor would exploit a tweet with a mea culpa flag like this on it. But I am not as imaginative as bad actors.

Re: De-escalating social media conflict

#126

Post by oehpr » Sun, Jul 05, 2020, 2:08 AM UTC

memexy wrote: post #113

> Reply to your tweet, posting "I was wrong" There is another option. Quote tweet and say "I was wrong". Quote tweeting is a concept extension mechanism but people mostly use it for dunking. If you make a mistake then quote tweeting is the best way to correct it because people can then search for "I was wrong" and see the wrong tweet in context.

I don't believe that would work, because links to the original tweet will still be broken, allowing people to say "X deleted their tweet" without saying that they reposted and said they were wrong.

Motivated actors loooooove this form of lying.

Re: De-escalating social media conflict

#127

Post by noobermin » Sun, Jul 05, 2020, 2:11 AM UTC

npunt wrote: post #105

noobermin wrote: post #57

I think this work is great and interesting, and I respect it. One of the things I feel like it lacks is a discussion of the broader context: we all sort of suspect that one of the reasons Twitter does not generally allow for de-escalation (like FB not censoring misleading news articles) is that such escalation is in fact part of their revenue stream (escalation=more engagement), so while I think ideas like this would…

I don't think twitter is not allowing de-escalation as much as they haven't figured out how to do it. The product itself hasn't changed much in a long time, and there are likely many internal reasons for that that have little to do with revenue. Meanwhile the FB misleading articles example is a much more nuanced problem that has less to do with engagement than simply being a political hot potato and a really challeng…

I should temper my above comment (my own sort of mea culpea :) ). You're right, given that I don't work for Twitter, I don't know whether there is an explicit desire not to reduce toxicity due to revenue concerns. I will say whether I am right or not, there at the very least does exist the incentive especially if they ignore the long term issues as you mention. Some evidence that at the very least they don't have an incentive to detoxify is that twitter has been toxic for years, almost a decade at this point. One of the most often cited essays on twitter toxicity from a leftist perspective was written in 2013[0] but the opening discussion sounds like it concern today.

All in all, I think you're quite correct about how the toxicity is a long term risk, but perhaps people there need to realize how much of a threat it is for the platform.

Finally, I think your article was really great, as I said before. I came with a negative perspective but you convinced me otherwise. I still believe that certain cultural and mindset issues need to be addressed that might be beyond design but I definitely feel you're right that the UX design could help reduce toxicity.

[0] https://www.opendemocracy.net/en/opendemocracyuk/exiting-vam...

Re: De-escalating social media conflict

#128

Post by nullc » Sun, Jul 05, 2020, 2:12 AM UTC

andrepd wrote: post #89

ssalazar wrote: post #86

Earlier quoted context omitted.

"Engagement" is the "paper clip maximizer"[1] of the modern software industry, or more generally an example of "instrumental convergence." Optimizing engagement without regard for negative externalities is a catastrophic ethical failure of the major players in the software industry, and we are finally seeing a backlash both from mainstream advertisers and widespread public outcry. [1] https://en.wikipedia.org/wiki/In…

Why should a for-profit enterprise have any kind of ethical consideration at all? It seems that's your mistake.

In a world of pure capitalism without ethical considerations at all... these sorts of issues can be "resolved" through assassination markets against the owners and operators of enterprises that become too great a public nuisance.

Is that what you want, or would you instead prefer that ethical considerations be back on the table?

Re: De-escalating social media conflict

#129

Post by nullc » Sun, Jul 05, 2020, 2:15 AM UTC

shadowgovt wrote: post #53

rowanG077 wrote: post #42

Earlier quoted context omitted.

Exactly. From what I have seen an apology only makes things worse.

A badly-executed apology divorced from change makes things worse.

No, any admission of fault or error makes things worse.

The public is utterly bombarded with claims that this or that person or group was wrong or did wrong with respect to something or another.

It's extremely difficult to sort out which of these allegations are true or false. Unless you admit fault, then its obvious to everyone.

Re: De-escalating social media conflict

#130

Post by ChrisMarshallNY » Sun, Jul 05, 2020, 2:17 AM UTC

hliyan wrote: post #118

ChrisMarshallNY wrote: post #111

Earlier quoted context omitted.

I was an engineer and a manager of software development for a very well-known, arch-conservative Japanese optics corporation. I worked there for just shy of 27 years. LOTS of 14-hour flights (shudder). But I am not Japanese, so it's quite possible they treated me differently from others. My experience was, they had a difficult time, trusting non-Japanese. They tended to pull me in, when dealing with westerners. It wa…

> I was in a "silo" of extremely honorable, incredibly high-quality people for a couple of decades Have you written more about your experiences during this period? I would love to read them!

Not especially. It is a corporation that is very protective of its privacy and secrets.

I tend not to name it, even though it's fairly easy to figure out.

I don't want to cause them any stress or harm. Even though I wish things had gone somewhat differently, I retain the greatest respect for the corporation, and its employees and officers.