Stefen Auris@pawb.socialM to Furry Technologists@pawb.social · 9 months agoResearchers Tested AI Watermarks—and Broke All of Them | WIREDwww.wired.comexternal-linkmessage-square2fedilinkarrow-up115arrow-down10cross-posted to: hackernews@lemmy.smeargle.fanshackernews@derp.foo
arrow-up115arrow-down1external-linkResearchers Tested AI Watermarks—and Broke All of Them | WIREDwww.wired.comStefen Auris@pawb.socialM to Furry Technologists@pawb.social · 9 months agomessage-square2fedilinkcross-posted to: hackernews@lemmy.smeargle.fanshackernews@derp.foo
minus-squarejsdzlinkfedilinkarrow-up8·9 months ago Watermarking at first sounds like a noble and promising solution Does it? Watermarking typically aims to be difficult to detect and hard to remove. To me it sounds more like the steganographic equivalent of malware.
minus-squarephonyphanty@pawb.sociallinkfedilinkarrow-up1·9 months agoHow so? Personally I don’t see AI watermarks as having the same adverse effect as malware.
Does it? Watermarking typically aims to be difficult to detect and hard to remove. To me it sounds more like the steganographic equivalent of malware.
How so? Personally I don’t see AI watermarks as having the same adverse effect as malware.