Comment on Eventually ChatGPT will eat itselfparentComments−curtisblaine3yIf ChatGPT is able to emit output that is watermarkedThen the watermark should be made public. We need a way to tell ai-generated content from human-generated content without actually having to read it.−sfg3yI wonder how difficult it will be to apply the filter to human generated content.
Comments
Then the watermark should be made public. We need a way to tell ai-generated content from human-generated content without actually having to read it.
I wonder how difficult it will be to apply the filter to human generated content.