Tim Davies argues that deepfakes have made it impossible to trust video or audio anymore. AI labs are adding watermarks to detect synthetic media, but ordinary people can't verify them, and adversaries will strip them anyway. Media trustworthiness is now fundamentally broken.
we can't trust anything that's put in front of us anymore, really, can we? And as consumers, we already can't trust anything we see or hear. We know that audio can be cloned. We know that videos can be created from anything. Essentially, you can't trust video anymore. You can't trust really anything that's being shared. And yet the AI labs are implementing fingerprinting, and even in the text that's output by these models, they're they're outputting in such a way that they can take it back and say, okay, this was our content. Yeah. But as general consumers, we don't really know. And there will always be someone somewhere that develops a solution that doesn't have those water markings in place. Yeah. So we can't trust anything that's put in front of us anymore, really, can we?
EP 60: Why Founders Become the Bottleneck — And How AI Changes Everything | Thea Brook
