I have to admit, I have been in the camp of wild optimism about artificial intelligence.
After reading this book, I have changed my mind. The points are very logical, easy to follow, and the analogies are a nice way to add color.
I won’t be taking orders from an ant or a lab rat, and likewise a Superintelligence will not be taking orders from me.
It’s not an exaggeration to say that this technology is more dangerous than the proliferation of nuclear weapons.
It will be very difficult to get international cooperation on this matter, but we have to try.