Once you've created an intelligence so smart, the real job of that intelligence is to protect itself from other intelligences becoming more intelligent than it. It's just kind of like human beings. The way you look at money or the way you look at the success of your child, you always want to make sure that as far as it gets, it can protect itself and continue forward. So I think any type of intelligence, no matter what it is, is going to have this very basic principle to protect the power that it has gained.

It amazes me that we spend 20% of the US budget on defense and far-off wars, and not on fighting cancer, disease, and aging.

My goal is to try to tell the public that America could use more science and technology in all aspects of its life.

In the future it's very possible you could have an artificial intelligence system that can run the country better than a human being. Because human beings are naturally selfish. Human beings are naturally after their own interests. We are geared towards pursuing our own desires, but oftentimes, those desires have contrasts to the benefit of society, at large, or against the benefit of the greater good. Whereas, if you have a machine, you will be able to program that machine to, hopefully, benefit the greatest good, and really go after that.

The American Dream has become a death sentence of drudgery, consumerism, and fatalism: a garage sale where the best of the human spirit is bartered away for comfort, obedience and trinkets. It's unequivocally absurd.

Frederick Nietzsche was important to me, in teaching that it's okay to strive to improve the human being.

I think we're already getting to a stage where the basic artificial intelligences are discovering moral systems. I think, in many ways, moral systems are simply things that we have programmed into ourselves, either through childhood or just through genetic, ingrained ideas. So the same thing applies when you talk about machines.