This may seem off topic but what do we do when bots catch up to our state of the art for keeping them out? Do we milk the concept of captchas until they turn i

This may seem off topic but what do we do when bots catch up to our state of the art for keeping them out? Do we milk the concept of captchas until they turn into some kind of Turing test? Can we even automate a Turing test in a way that machines can't bypass? Do we basically ditch the concept of privacy and require some kind of systematic provable identity in most places? Or do we accept the fact that consensus can be faked by bots and try to mitigate the noise by removing the spam as it appears?