“You basically make machines that have human characteristics, intelligence is one of those features,” Hoffman said.

We use AI all the time, and for many of us, and we do not even know that. Your TV has, your coffee machine has, and your car is using it. Think about anti-lock brakes or towing. IPhone Siri and Alexa from Amazon are also obvious examples. Everything is called AI Narrow because they are really good at task. But if you have asked your anti-skid brakes to solve a mathematical problem, you will have problems.

“I think there is something we are worried about, that could be the ability to communicate and learn,” Hoffman said.
Do not you understand that seriously? Perhaps because he still feels so fictional, and is visible in many movies, from iRobot to war games. Super intelligence is fiction for now, but again, the space developed until we were.

“We develop atomic energy with the highest intentions and it turns out that you can make a bomb,” Hoffman said.
Now about bad results for us. Judgment day AI, if you will. It could be something as simple as misinterpretation of a computer, like booting robots in iRobot. Instructions AI are guarding people, central spirit and believing that the only way to do this is basically to keep us in prison so nobody goes.

Or, if super intelligence decides that people are no longer needed, and we are just useless unpleasantness. Well, what can we do now? Well, not much, except knowing it.
The founder of Spacex and Tesla, Elon Musk, has created Open AI, a non-profit corporation that wants to improve AI, announcing all its online results for everyone.

Google has invested billions of dollars to improve computer power in a safe way

Since the strength of computing continues to accelerate, the best way to hope is that the progress of AI looks more like Jetson and less on Terminator.