Dangerous approach.
Bad actors will not watermark their output, or remove the watermark. All watermarking does is lend credibility to misinformation. It’s literally worse than nothing.
Dangerous approach.
Bad actors will not watermark their output, or remove the watermark. All watermarking does is lend credibility to misinformation. It’s literally worse than nothing.
redcalcium@lemmy.institute 9 months ago
It won’t stop bad actors, but it’ll allow AI companies to cover their asses to avoid being blamed for misuse of their tech, which is why I think most AI companies will use it soon.
General_Effort@lemmy.world 9 months ago
I’m sure you’re right on that. I think it’s already done.