We have no idea how to have a program detect AI-written text in a useful way.
This approach seems very doable:
The tl;dr here is that you can make an AI that fingerprints its output in a way that can be detected later, but this doesn’t generalize to output from other AIs.
Why couldn’t OpenAI just keep a database of all their outputs?
This approach seems very doable:
The tl;dr here is that you can make an AI that fingerprints its output in a way that can be detected later, but this doesn’t generalize to output from other AIs.
Why couldn’t OpenAI just keep a database of all their outputs?