Reading Superintelligence by Nick Bostrom. Either we’re already living in an simulation or it’s all bullshit. I’m unsure which it is. The book is mostly a philosophical overview of the point at which machine intelligence becomes\u00a0 “conscious” and begins to optimize for whatever goal it has determined. This is the crux of the problem – how can humans seed machine intelligence so that the goal of the AI is beneficial … or in the least doesn’t optimize for a goal that destroys humanity. It’s not a science fiction book. Consider the IQ difference between the village idiot and Einstein. The difference is nominal. Now consider a machine with an intelligence that is 1000x Einstein. What does that even mean? What would it want to achieve? Terrifying.<\/p>\n
When I’m feeling the weight of the world or push myself up a long climb while running, I growl. It’s a primitive queue to try harder. And I do.<\/p>\n