Some new work from our lab relevant to those thinking about AI harms in the here and now! If you're worried about manipulated or synthetic image/video media spreading misinformation on social media, one approach that may get us out of the losing battle of detection is provenance. See thread: @kjfeng@hci.social