Conversation

Notices

1. Embed this notice
   Judge Dread (judgedread@poa.st)'s status on Monday, 11-Mar-2024 15:35:38 JST Judge Dread
   I consider that 'study' of 'violent' memes from Parler of all places a smoking gun for the Gab meme ban. I've been predicting a big shuttening in the rampup to the election and the rampup began March 1. That's why Torba rushed to introduce the 'feature' - he had his orders.
   
   Oh, and of course from the perspective of the enemy Parler is the source of most current active Gab users, so they might consider Gab a good test bed for their countermeasures.
   • Embed this notice
     Judge Dread (judgedread@poa.st)'s status on Monday, 11-Mar-2024 15:35:36 JST Judge Dread

     in reply to

     • p
     @p There was no more to the meme, but the meme had enough info to find the following. For those reading along at home be careful about opening PDFs, last I checked they could still contain tracking pixels, only read with internet off or with a nice stronk VPN.
     
     One twist is that the report is 11 months old, but they chose to resurrect it just before Torba's meme ban. My guess is they have a whole quiver of these bullshit reports and studies to signal boost in series to justify what's coming. Now, the data and links.
     
     This is the article that revived this stale study for use as weapon:
     
     polcomm.northwestern.edu/ccpp-report-highlights-the-danger-from-violent-memes-in-2024-u-s-election/
     
     CCPP Report Highlights the Danger from Violent Memes in 2024 U.S. Election
     
     Posted February 23, 2024 by Ani Feinberg
     
     Memes have become a critical communication tool for partisan and extremist groups. CCPP scholars took a deep dive into the content of memes on alternative (or fringe) platforms, and what they found is just as concerning as it is fascinating. There is no doubt that 2024 will be saturated with online political content, but some of the most dangerous content should be expected to come in “meme” form.
     
     Sponsored by a grant from the Global Network on Extremism and Technology (GNET), a CCPP team employed state-of-the-art deep learning image and visual rhetorical analysis to analyze memes posted on far-right platform Parler. Detailed in the CCPP Report, the team categorized memes into themes of gender, race, partisanship and violence. Researchers then measured the transmission rates of memes associated with each theme, finding, among many other patterns, that violent memes are on the rise. It was also noted that memes with high engagement levels were often branded by a group emblem or logo, increasing their credibility within public discourse.
     
     Under the guise of an unassuming image or inside joke, memes tend to hide their political intent and are more able to escape algorithmic content moderation or filtering. However, the report highlights how moderating memes is critically important for limiting online content that promotes political violence. The methodology and findings of the report provide a working model for meme content filtering that can help platforms identify and filter memes with extremist content.
     
     That article links to this:
     
     kclpure.kcl.ac.uk/portal/en/publications/a-picture-is-worth-a-thousand-swords-classi%EF%AC%81cation-and-di%EF%AC%80usion-o
     
     There you find a direct link to the PDF:
     
     kclpure.kcl.ac.uk/portal/files/202622942/GNET_33_Meme_Classification_on_Partisan_Platform_web.pdf
     
     A Picture is Worth a Thousand (S)words: Classification and Diffusion of Memes on a Partisan Media Platform
     
     Ayse D. Lokmanoglu, Mowafak Allaham, Rod Abhari, Chloe Mortenson , Esteban Villa-Turek, GNET (Global Network on Extremism & Technology)
     
     Original language
     English
     Publisher
     Global Network on Extremism and Technology
     Number of pages
     48
     Publication status
     Published - 29 Mar 2023
   • Embed this notice
     p (p@shitposter.club)'s status on Monday, 11-Mar-2024 15:35:37 JST p

     in reply to
     @judgedread What was the study? Have a link?
     
     (Related, I think you saw the MIT Sloane guy that was scraping a few months back, it looked like an asshole rather than something sinister, but it turned out to be sinister, he showed up in the NSF presentations: https://media.freespeechextremist.com/rvl/full/835374c1bfa10895663d4d1c94500049823ea928fb7e9c47b01a6b7f8f07c091 .)
   • Embed this notice
     p (p@shitposter.club)'s status on Monday, 11-Mar-2024 15:35:49 JST p

     in reply to

     • lpheathen2 :verified:
     @lpheathen2 @judgedread Well, yeah, but I'm unreasonably paranoid and I still think "Don't click a link" is a bit much. It's like the threat model is too vague and ill-defined, and when the alarms go off over everything, it's easy for the actual threat to sneak past.
     ✙ dcc :pedomustdie: :phear_slackware: likes this.
   • Embed this notice
     lpheathen2 :verified: (lpheathen2@poa.st)'s status on Monday, 11-Mar-2024 15:35:50 JST lpheathen2 :verified:

     in reply to

     • p
     @p @judgedread understood
     I guess I assumed that most people would take online anonymity more seriously as the threat of doxing can ruin someone’s life.
     i don’t see it as paranoid so much as a safety that prevents threats and also assumed that most on here (poast) want to hide their free speach activities for the same reason.
     then again maybe I’m just paranoid after all
     my opinion doesn’t really matter much as I’m not a huge account on the netz, i just believe in personal safety
   • Embed this notice
     lpheathen2 :verified: (lpheathen2@poa.st)'s status on Monday, 11-Mar-2024 15:35:51 JST lpheathen2 :verified:

     in reply to

     • p
     @judgedread @p I truly thought most people on here had some form of a sandbox set up for safe viewing of threat material from threat vectors that monitor for IoT. It’s always nice seeing a warning ( don’t click links) (don’t open rando pdf’s)
   • Embed this notice
     p (p@shitposter.club)'s status on Monday, 11-Mar-2024 15:35:51 JST p

     in reply to

     • lpheathen2 :verified:
     @lpheathen2 @judgedread
     
     > I truly thought most people on here had some form of a sandbox set up for safe viewing of threat material from threat vectors that monitor for IoT.
     
     The chudbuds.lol admin got her tax returns and divorce filings and child support payments and everything down to photos from the hospital of her new husband's dick injury, all posted online. She ran an instance with 2100 active users ( http://demo.fedilist.com/instance/chudbuds.lol ) on it, and she used the same Windows computer to administer that machine, handle business, do her Twitch streaming, and also install RAT tools disguised as Minecraft mods. I am not sure what qualifies as "most people" but I think people that are remotely careful about anything are vanishingly rare. Most people don't really have a reason (although in Claire's case, installing a random executable on the computer she used for everything while also acting as an admin for a very noisy instance is possibly the biggest mismatch between how careful someone should be and how careful they are).
     
     > ( don’t click links) (don’t open rando pdf’s)
     
     In thus case, it's a known domain belonging to a regular site and it's an academic PDF and we're not at "don't click links" threat-levels yet. That's Gab levels of of "What if they find my IP address?" paranoia. It's just shitty people publishing an academic dry-hump. It's not a thing to worry about. Tor's sufficient for now.
     ✙ dcc :pedomustdie: :phear_slackware: likes this.