People are now using Grok to «verify» videos of attacks, and as expected, each time Grok delivers different, mutually exclusive but plausibly sounding versions of the «facts.»
And yes, this is essentially the same AI model now being used by the US military in classified operations
#WildForensis
Screenshot of Tweet with 12 Million views: »Breaking Report: Harris Campaign BUSTED using a FAKE crowd photo at campaign rally stop«, illustrated with the detail of a reflection
Nevertheless, people began looking for alleged traces of AI manipulation in the official photos of Harris's recent campaign rally stop, highlighting supposedly telling details while ignoring the widely available video footage of the very same event. This is what I call #WildForensis
2/3
Like #WildForensis, there is now a #WildIconology. Everyone online is now an iconologist. Everyone knows: this is an icon. Iwo Jima here, 9/11 there, the comparisons are on standby. The image in question is always already fully interpreted and yet reproduced again and again
