207: Here's The In-Depth Treatment
...........................
If you want the quick take, here's the link. If you want the juicy, deep-in-the-weeds approach, keep reading.
There's perhaps nothing more disappointing than seeing the birth of our social media world--technology's promised land, uniting us all in one band of equal compatriots--ruined by losers needing to put others down in order to feel themselves. From unsportsmanlike behavior, to hypocrisy, to creepiness, all the way to out-and-out harassment, the usual depressing verdict is that there's no way to stop all but the worst foul play without violating free speech in the process.
And yet, engineering social media to take care of trolls and other assorted haters is almost certainly doable. As noted in the quick take post linked to above, Twitter is considering ways to accomplish this using algorithms, as well as the usual human admins, to catch problems. One much simpler solution briefly outlined in the above post is to add an 'un-uh' button, what I call a 'ding' button, that would be used on one's own posts (in the comments section), on Direct Messages (DMs), and on Mentions (links embedded in other peoples' posts that include you in a conversation). Retweets (where someone republishes your tweet, adding commentary) would, on the other hand, be fair game (Twitter has ways to 'mute' such encounters).
Most importantly, accounts that accumulate dings would become Red Flags for Twitter's screening process. Plus, I propose an automatic closing down of any account with a ding and no 'likes'.
And, since the harassers are simply going to sign up using a different account, I suggest that Twitter integrate Red Flags into their sign up process to weed out the haters.
Now, let's ask the questions surrounding feasibility, examining each question with common sense, and hopefully, at some point, there'll be expert opinion brought to bear here as well.
1. Stopping Repeat Offenders: Assuming a Red Flag warning (an account with three strikes--three of our dings--against it), would Twitter be able to block the user during an attempted second sign-up using a different account?
Non-Expert Answer: A hard core hater is going to have multiple devices, each with multiple accounts, so each account would be used at least once to harass. But, for the half-hearted harasser with just one device and little time to go to the trouble of setting up multiple accounts, one would think that a screening could be done, comparing a user's device with a database of terminated Red Flag accounts. One would think that this would effectively weed out a large percent of the problem.
If an immediate shut down of a dinged account could take down all accounts associated with that device/user, this would eliminate even more, including those setting up harassing accounts on the side, in addition to a normal Twitter presence. Actually, dissuading, in the form of a new rules rollout would probably be all that would be necessary in many cases.
2. Accidents: Could accidental 'hit-the-wrong-button' mistakes, misunderstandings, and sabotage possibly occur? For example, a device is stolen, hate messages posted, and a lifetime account is then in jeopardy. Or, a ding is hastily doled out due to a modest disagreement. Or, a button is hit accidentally.
Non-Expert Answer: If the 'ding' button triggered a dialog box that required 2-step activation, this would remove most accidents. Plus, dinging could conceivably be followed with an undo? question that was always actionable.
The solution to misunderstandings is also fairly straightforward: the offending comment, DM, or Mention, and its context, would be publicly accessible within a user's list of dings (just as 'likes' are easily looked at). So, a misunderstanding could be annoying, but easily explained. Plus, ideally, Twitter would deputize volunteers (those with a reputation for civility), to expunge dings that were due to obvious errors and minor misunderstandings.
And what about sabotage, when a device is stolen, and someone else purposefully accumulates dings? Allowing for accounts to be reinstated, once such an occurrence is identified and brought to Twitter's attention, would likely be sufficient.
3. Who Has Time? Who'd have time to patrol their account? And would this do any good
if harassers coordinated their attacks?
Non-Expert Answer: I'd have time, though not everyone would. But when you do have time to double-check, you'd have the power to shut the nastiness down. Psychologically, this would be considerable consolation in itself. And if Twitter could screen its sign-up process using a Red Flag database, harassment would, ideally, be a quick ticket to banishment, eventually removing all but the most persistent harasser.
As for coordinated attacks, if all fraternal accounts associated with a particular internet address were knocked out due to a three strikes policy, the problem would act to remove itself from the platform. So, short term, a nuisance; long term, good riddance.
4. Include Retweets? This is probably where ideas like this have failed, when considered by Twitter. Though a retweet is your tweet being repurposed; including retweets in this construct would be a bridge too far. Right?
Non-Expert Answer: Right. If one had to worry about retweeting the latest content, Twitter would soon be a ghost town. Leaving retweets alone means there's still room for a Wild West atmosphere--within Twitter's rules of conduct. But, importantly, merely retweeting, and adding a harassing remark, isn't going to break through to a larger audience (assuming a tweak or two to Twitter's algorithm) if the only users reading it are a harasser's fellow harassers; this, as opposed to comments and mentions, which expose harassers to everyone else. Plus, a retweet could always be reported to Twitter if it crossed a line.
Overview
So, unless my amateur's understanding is mistaken, the seemingly impossible task of monitoring billions of twitter messages in real time could be easily done simply by adding a 'ding' button, and letting users do the work themselves.
There'd have to be a gradual rollout, so that any new rules and their consequences were understood. Perhaps weed out the worst offenders, then ratchet things down to a mere three strikes with reinstatement possible in some circumstances. Maybe send a warning message after any Red Flag occurrence.
What makes the 'ding' button idea especially interesting for Twitter, if one were to consider its use on other platforms, is that Twitter posts are open to the world (unless one uses Lists, as I understand it), compared to a platform like Facebook, where content is often posted to 'just Friends', or even more intimate groups. If this is true, Twitter has both the bigger problem to fix (harassers can't get to much Facebook material), but also much more to gain. If it were to become uniformly civil, Twitter would be the broader, more exciting stage on which to tread.
No comments:
Post a Comment