Re:Researchers Tested AI Watermarks–and Broke All of Them
by ryukoposting on Friday October 06, 2023 @07:22PM
(#37797580)
I was thinking the same thing- we've had digital signing algorithms for decades, and those seem to work fine enough. There's a healthy distrust of cert authorities, sure, but it still works.
[ Parent
| on HN ]
Re:Researchers Tested AI Watermarks–and Broke All of Them
by p1necone on Friday October 06, 2023 @08:24PM
(#37797958)
The signature only proves that the website/Bill the photographer was involved in the chain sending the website/photograph to you, not anything about the content itself.
Unless you have reason to trust Bill himself you can't trust that he actually took the photo, or that it isn't ai generated. Although knowing that Bill isn't tech savvy enough to do those things might be enough.
Re:Researchers Tested AI Watermarks–and Broke All of Them
by jacobr1 on Friday October 06, 2023 @10:59PM
(#37798639)
That ... might be all we get though. If anybody can produce artificial, but realistic content that is indistinguishable from something the real thing, then all we might have is our willingness to trust any given originator/distributor of the content.
Narrowing down the problem to "Do we trust Bill" is at least something we can attempt to address. This also eliminates some sources where we can more easily discount things. If the source is 4chan, maybe I don't need much more information to make my repetitional assessment.