AI & ROBOTICS
SYNOPSIS OF COAST TO COAST AM SHOW
In the first half of Tuesday's show, aerospace and defense systems developer Sir Charles Shults discussed the rapid advances in robotics and artificial intelligence (AI) and why some people fear they will dominate the available jobs for humans, as well as eventually gain an upper hand over humans. "If you build a machine that's smart enough to clean up after you, maybe it'll realize it doesn't want to do those jobs just as you don't," he remarked, adding that the question becomes, can we throttle back the self-interest of the machine in its design? And, "if you have a machine that's perhaps smarter than a human being, how do you know whose interests it's acting in?"
We are so close to opening a Pandora's Box, with an explosion of AI coming our way in the next decade or so, Shults commented. It actually won't be that difficult to produce emotions in the machines, and there's going to be many different types of AI, and "we're not going to know which ones are potentially dangerous and which ones are very beneficial to us until we actually see them in practice," he stated. Because of this, Shults suggested "sandbox systems" or virtual worlds where the AI can be tested before being released into the real world. Further, he continued, everyone who works on artificial intelligence should have to take a course on morality, and unintended consequences, in order to think through how these machines will interact with humanity.
LIKE FRANKENSTEIN THE CREATORS FALLS IN LOVE WITH THE CREATED AND THE HELL WITH THE CONSEQUENCES
LIKE THE ATOM BOMB
SYNOPSIS OF COAST TO COAST AM SHOW
In the first half of Tuesday's show, aerospace and defense systems developer Sir Charles Shults discussed the rapid advances in robotics and artificial intelligence (AI) and why some people fear they will dominate the available jobs for humans, as well as eventually gain an upper hand over humans. "If you build a machine that's smart enough to clean up after you, maybe it'll realize it doesn't want to do those jobs just as you don't," he remarked, adding that the question becomes, can we throttle back the self-interest of the machine in its design? And, "if you have a machine that's perhaps smarter than a human being, how do you know whose interests it's acting in?"
We are so close to opening a Pandora's Box, with an explosion of AI coming our way in the next decade or so, Shults commented. It actually won't be that difficult to produce emotions in the machines, and there's going to be many different types of AI, and "we're not going to know which ones are potentially dangerous and which ones are very beneficial to us until we actually see them in practice," he stated. Because of this, Shults suggested "sandbox systems" or virtual worlds where the AI can be tested before being released into the real world. Further, he continued, everyone who works on artificial intelligence should have to take a course on morality, and unintended consequences, in order to think through how these machines will interact with humanity.
LIKE FRANKENSTEIN THE CREATORS FALLS IN LOVE WITH THE CREATED AND THE HELL WITH THE CONSEQUENCES
LIKE THE ATOM BOMB