Imagine how Eric’s story might be different if we are successful. As a novice user of Twitter, Eric would have a medium or low truthfulness score — like a first-time ridesharing driver or eBay seller. His tweet would have his score attached to it and so would the image he attached. You can imagine a…
Interesting concept, though I assume you’d need two scores; one for the tweet, one for the link posted (if any). This is because fake news would otherwise be seen as ‘real’ if a ‘reliable’ source gets fooled by it, and vice versa.
Other minor issue is the question of ‘who’ decides whether someone is ‘truthful’, and ‘how’ that works. Do you assume someone who’s popular or active o Twitter is more reliable? Might help for a few new/blatant troll accounts, but it’d also have problems with popular but controversial figures like say, Donald Trump.
That said, having the ability to mark a tweet as false in retrospect would be incredibly useful for everyone (especially journalists).