3 ms·
Seems like Meta could and should detect the image-of-article workaround. It's easy enough to use OCR and then match to a known news source.
by bitshiftfaced 2y ago
Seems like Meta could and should detect the image-of-article workaround. It's easy enough to use OCR and then match to a known news source.
- beej71 2y agoAt FB scale, this seems like a non-trivial computational expense, meaning they'll do it if they're forced to.
- bobthepanda 2y agoI would imagine they have similar systems in place from back when ISIS used Facebook Live to stream beheadings.
- bitshiftfaced 2y agoFair enough. I was thinking maybe long term it could save compute and business complexity, since the idea of undoing the law won't gain momentum if people just use workarounds.
- da768 2y agoPeople share links to linktr.ee pages and said page has a link to the news article. OCR would be no use unless Facebook started crawling links