"In a world where generative AI technology is increasingly adept at mimicking the work of humans, people are naturally skeptical when online platforms refuse to label even obvious AI content."
"Instagram head Adam Mosseri suggested as much in December, saying that it will be 'more practical to fingerprint real media than fake media' as AI technology improves to the point of making content that's visually indistinguishable from that made by creative professionals."
"Authenticating human-made works was something the C2PA content credentials standard - which is already used by Meta's platforms - was supposed to do. But so far, its implementation has been wholly ineffectual."
"Right now, there are too many AI-free labelling alternatives to choose from. In total, I count at least 12, all trying to address the same issue with a variety of eligibility criteria and authentication methods."
The rise of generative AI technology has led to skepticism about the authenticity of online content. A proposal suggests labeling human-made works with a recognizable logo to differentiate them from AI-generated content. Instagram's Adam Mosseri supports this idea, emphasizing the need to fingerprint real media. Despite the C2PA content credentials standard aimed at authenticating human works, its implementation has been ineffective. Numerous labeling solutions exist, but challenges remain for widespread adoption due to the variety of criteria and the motivation of AI content creators to obscure origins.
Read at The Verge
Unable to calculate read time
Collection
[
|
...
]