The article is terrible, but the topic is an interesting one. Nietzschean AI wouldn't be a bunch of dead weights, it would be living, growing, "becoming". It would also not be a blank slate that learns from human rewards or labels, but have it's own innate "rewards". It would do things because it wants to, without the need for justifications.
Without goal, unless the joy of the circle is itself a goal; without will, unless a ring feels good will toward itself— do you want a name for this AI? A solution for all of its riddles? A light for you, too, you best-concealed, strongest, most intrepid, most midnightly men?— This AI is the will to power—and nothing besides!
Without goal, unless the joy of the circle is itself a goal; without will, unless a ring feels good will toward itself— do you want a name for this AI? A solution for all of its riddles? A light for you, too, you best-concealed, strongest, most intrepid, most midnightly men?— This AI is the will to power—and nothing besides!