We regularly hear people concerned about the hypothetical existential threat of a paperclip maximizer AI being invented in the future, yet this concern reliably ignores the AI overlords we already have: paperclip^Wprofit maximizing corporations. The VM for this type of AI has an incredibly slow clock rate and an extreme CISC ISA that often modifies itself in the RTC interrupt. The AIs are not a future threat; they already enslaved us.
The problem is our unregulated capitalism. Until that is addressed[2] we are merely debating which profit-maximizer AI we want to serve.
(my thanks to Charles Stross for the maximizer-AI/corporation metaphor[1])
[1] https://media.ccc.de/v/34c3-9270-dude_you_broke_the_future
[2] I suggest heavy regulation to protect the useful parts of capitalism. Left alone, capitalism destroys the very markets it needs to survive.