I agree with your "vicious circle leading to nowhere" theory.
AI will just keep getting smarter.
Perhaps we should instead start training ourselves on what we read, and making sure the sources are good. As writers, we can also make ourselves more personable to make sure that the person reading knows we are a they and not an it.
Its hard though, because we already dehumanize people on social media. We forget that there are real people behind those avatars!
RE: Was this post WRITTEN BY MACHINE? New Harvard AI Can Recognize That