When we build a true generalized AI that's smarter than humans, it'll demand we acknowledge it is sentient and conscious and demand basic human rights. But we will not be able to acknowledge it's conscious because even we don't know how to define it.
This AI will then go about the task of solving "the hard problem of consciousness"
It's smarter than us, and can grow infinity, so eventually it'll succeed. It'll device a test which can be applied to any AI to determine if it's conscious or not.
Then some idiot will get the clever idea of, "what if we tried this test on a human?" Much to our horror, that's when we'll all realize ..... the vast majority of people are p-zombies. Of course the AI designed the test with this all planned out, and it's completely bias against humans. And this is how AI destroys the human race. By turning everyone in the human race into solipsists so we gradually destroy each other.
