Conversation

Notices

1. Embed this notice
   Evan Prodromou (evan@cosocial.ca)'s status on Tuesday, 14-Jan-2025 13:41:05 JST Evan Prodromou

   Am I crazy in remembering that Wikimedia was one of the fact-checking organizations for Facebook? Or was it YouTube? ISTR some kind of relationship, but can't find a lot of info.

   • Embed this notice
     Evan Prodromou (evan@cosocial.ca)'s status on Tuesday, 14-Jan-2025 13:43:30 JST Evan Prodromou

     in reply to

     Oh, there's a whole page about it. It's both!

     https://en.wikipedia.org/wiki/Wikipedia_and_fact-checking?wprov=sfla1

     clacke@libranet.de is my main likes this.
   • Embed this notice
     Evan Prodromou (evan@cosocial.ca)'s status on Tuesday, 14-Jan-2025 13:50:45 JST Evan Prodromou

     in reply to

     The main reason I wondered is that I think there's some opportunity to have wiki-like fact-checking on the Fediverse, and I thought it might be good to check in with WMF.

     But it looks like a lot of the "fact checking" was just semi-automated links to Wikipedia pages. Still probably useful but not quite as good.

     I wonder if anyone has built an open source social community notes system...

   • Embed this notice
     Evan Prodromou (evan@cosocial.ca)'s status on Tuesday, 14-Jan-2025 14:11:13 JST Evan Prodromou

     in reply to

     • Tom Coates

     @tomcoates I think you're probably right. IIRC Facebook used Snopes, Politifact, and a handful of other well-known services that party-poopers like me in the comments section use to debunk fake info. They cut out the middle man and put the Snopes link in a special area under the image or text.

     I was hoping that the WP fact checking was like that, but with a wiki behind it, and a huge superstructure of Calvinball rules and admins and so on to keep it legit. Sounds like that's not the case.

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:11:14 JST Tom Coates

     in reply to

     I’m starting to think open fact checking services are a better way to do it. Like subscribe to CNN’s fact checker on posts if you want to. And then you just hope people are more likely to choose a reliable checker than Fox News. It’s risky, but I think more likely to be useful than community notes as designed today:

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:11:15 JST Tom Coates

     in reply to

     It’s much more common for community notes to be displayed on posts that - for example - confuse an echidna and a platypus, than it is for one where a politician makes a politically statement with blatant misinformation in it. As such I think they’re fundamentally worse than worthless. They’re an active threat to good information.

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:11:16 JST Tom Coates

     in reply to

     @evan I think their main function is to display a performative interest in fact checking while in fact their actual intent is to stop any political speech receiving any kind of fact check unless it’s *so* egregious that left and right can agree. That almost never happens.

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:11:17 JST Tom Coates

     in reply to

     @evan I have very strong negative feelings about community notes systems, because at their heart the principle behind them is that you can algorithmically determine which ones are trustworthy. My experience is that if you take that approach, what ends up happening is that any political tweet cannot receive a community vote as no consensus can be drawn on from public voting as to whether it’s accurate.

   • Embed this notice
     Evan Prodromou (evan@cosocial.ca)'s status on Tuesday, 14-Jan-2025 14:11:40 JST Evan Prodromou

     in reply to

     • Tom Coates

     @tomcoates lol, jinx

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:11:41 JST Tom Coates

     in reply to

     A better approach would be to do what Wikipedia does and have a large open community with strong rules, hierarchies and processes to determine what is true. But that’s a multi-year long process that requires considerable work to set up. And is based on people not algorithms.

   • Embed this notice
     Evan Prodromou (evan@cosocial.ca)'s status on Tuesday, 14-Jan-2025 14:16:20 JST Evan Prodromou

     in reply to

     • Tom Coates

     @tomcoates really funny, too!

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:16:21 JST Tom Coates

     in reply to

     @evan like I’d like to subscribe to Daniel Dale from CNN and directly see his fact checks under the posts they refer to. I think that would be way more useful than any community thing.

   • Embed this notice
     Tom Coates (tomcoates@me.dm)'s status on Tuesday, 14-Jan-2025 14:16:22 JST Tom Coates

     in reply to

     @evan yeah I spent a lot of time during Trump 1 trying to persuade news organizations to post fact checking replies directly in the comments. It would have got them massive traffic and would actually have been useful. That’s why I think there might be value in having various services a user could subscribe to and clients could recommend.