Connect with us


Meta Expands ‘Redirect Initiative’ Which Offers Alternative Information Sources on Potentially Harmful Searches



Meta Expands 'Redirect Initiative' Which Offers Alternative Information Sources on Potentially Harmful Searches

Meta has announced a new expansion of its efforts to stop the potential radicalization of people via its apps, with its ‘Redirect Initiative’ being extended to two more regions.

As explained by Meta:

“We’re [expanding] one of our flagship counterspeech initiatives, The Redirect Initiative, into Pakistan and the UK. Now when someone in Pakistan or the UK searches on Facebook using words associated with organized hate or violent extremism, the top search result will be a link to resources and support for how to leave violence and extremism behind.”

Which sounds pretty simple – a little too simple to have any real impact. Right?

Well, not entirely. According to Meta, the Redirect Initiative, which is already active in Australia, the US, Germany and Indonesia, has proven successful in disrupting the connection between users and violent, extreme content.

Though not massively so. Meta’s contention here is based on this report, which shows that of 57,523 searches for such content in the app, over a three-month period between November 2019 and March 2020, 2,288 users clicked on its alternate, credible sources, which then lead to 25 of them seeking support from program partners.

That’s not a great success rate, but I guess, in Meta’s view, as well as those working with the company on the initiative, any disruption in this chain is significant, while the 2k clicks on these alternate links also show at least some hesitancy in searchers, which could have extended deterring impacts.


And it might also be just the start, with Meta further noting that academic insight has shown that countering such actions with credible alternatives is the best way to combat radicalization, and essentially stop users tumbling down dangerous online rabbit holes.

See also  LinkedIn Provides Notes on Effective Campaign Measurement [Infographic]

“Research shows that challenging violent, extremist views with positive counterspeech is one of the most effective ways to combat hate and extremism. Counterspeech involves responding to hateful, extremist narratives with empathy and alternative perspectives, rather than shutting these conversations down. Counterspeech is only effective if it comes from credible voices, so we’ve partnered with various NGOs and community groups on multiple counterspeech initiatives around the world including community programs, toolkits and research.”

It’s an interesting initiative. Not an overly successful one right now, but as noted, there will be lessons learned, and progress made, which could see it become a more significant, influential factor in reducing the pathways to more extreme responses over time.

Meta’s also working to improve its detection measures, which involves working with partner organizations and platforms to identify key trends, and in combination, all of these elements may end up having a more significant impact.

We now know, or at least have a fair understanding, of the role that social platforms can play in this chain, and if Meta can work with experts to develop better processes, then it is worth experimenting with such, even if the initial numbers aren’t wholly convincing.

Just one user deterred may well be enough, while every person who goes on to seek help can also provide further insight into exactly how this process evolves, from a psychological standpoint, which could glean key breakthroughs for expanded initiatives.

Source link



Elon Musk’s Team Asks for More Data to Complete Assessment of Twitter Bots



Elon Musk's Team Asks for More Data to Complete Assessment of Twitter Bots

Okay, let’s just check in on the latest with the Twitter/Elon Musk takeover saga, and where things are placed to close out the week.

According to the latest reports, Musk’s team recently asked Twitter for more tweet info, in order to help it make an accurate assessment of bot activity in the app. This comes after Musk questioned Twitter’s claim that bots and fake accounts make up only 5% of its active user base, and said that his Twitter takeover deal could not go ahead unless Twitter could produce more evidence to support this figure.

Which Twitter did, by providing Musk with access to its ‘full firehose’ of tweets over a given period, which it shared with Musk’s team back on June 8th. Musk’s group has now had that data for a couple of weeks, but this week, it said that this info is not enough to go on, and that it needs even more insight from Twitter to make its judgment.

And after initially resisting calls for more data access, Twitter has now reportedly relented and handed over more tweet data access to Musk’s team.

Which may or may not be a concern, depending on how you see it.

In its initial data dump, Twitter reportedly gave Musk’s team info on:

  • Total user tweets (within a given time period)
  • Data on which devices were used

As noted, Musk’s team says that this has not provided it with the insight that it needs to conduct an accurate analysis of potential bot activity, so Twitter has now provided Musk with more ‘real-time API data’.

See also  Google Search Offers Carousel In Search Result Snippets

It’s not clear whether that means that Twitter has provided everything that its API systems can provide, but that could mean that Musk’s team can now access:

  • Real-time info on tweet text and visual elements/attachments
  • Data on retweets, replies, and quote Tweets for each
  • Data on tweet author, mentioned users, tagged locations, hashtag and cashtag symbols, etc
  • Date, time, location, device info

That should satisfy any analytical needs to uncover potential bot trends, and get a better handle on Twitter’s bot problem, though it also means that Musk has all your tweet info – which, again, it’s worth noting, Twitter up till now had been hesitant to provide.

I’m sure it’s fine. Musk’s team is beholden to disclosure laws around such, so it’s not like they can do anything much with that info anyway, in a legal sense. But the idea that the sometimes erratic Elon Musk now has all the tweets could be a little concerning for some.

But Twitter likely had to provide what it can, and if Musk is going to become CEO of the app soon anyway, he’s going to have access to all of that data either way.

But still, given Musk and Co’s past history of undermining and attacking critics, sacking trouble maker employees and digging up potential dirt on rivals, it sits a little uneasy.

Should be fine. No problems – no need to go deleting all your DMs (which are likely not included in the data that Twitter has provided at this stage).

According to reports, Musk’s team says that it now has the info it needs to make its assessment of bot activity, which should see the deal move forward (or not) sometime soon.

See also  Facebook Provides Audience Connection Tips in Latest 'Social Skills' Video

Of course, no one knows what exactly is going to happen next, and whether Musk’s team will look to renegotiate, or even back out of the deal entirely as a result of its bot analysis. But it does seem like, one way or another, Musk will be forced to go ahead with the $44 billion transaction, with Twitter’s past bot reporting methodology already accepted by the SEC, giving it legal grounding to argue that it’s acted in good faith, regardless of what Musk’s team finds.

The next steps then, according to Musk, would be securing debt financing and gaining Twitter shareholder approval, clearing the last hurdles for Musk to change the app’s name to ‘Telsla Social’, and add a million references to ‘420’ into the platforms various terms and conditions.

Because of the memes, because weed jokes are still funny to the richest man in the world – because he vacillates between inspired genius and a massive nerd who now gets to play out some fantasy of being cool.


Or something. Who knows what goes on in Elon Musk’s head – which is also why most are hesitant to bet against him, as nobody knows if and how he might be able to fix Twitter, and whether this is a great investment or a massive disaster.

It seems like we may soon find out. Maybe. Who knows. Either way, the memes should be great.

Source link

Continue Reading

Subscribe To our Newsletter
We promise not to spam you. Unsubscribe at any time.
Invalid email address