Loading...

He had three lectures at the U of T University of Toronto of this year that examined AI risk, behavior, maintaining control with self-improving AI as well as how scientific oversight shared safety standards and international cooperation could help deal with the risk. And if you're following a progress of AI, it wasn't too long ago where it was simply just trying to construct some fluid sentences. But more recently, it solves some of the world's most complex problems. So it progressed pretty quickly. And as AI rapidly develops, future fears of AI agents developing new weapons like autonomous car, robot, AI, taking full of curable infrastructure and even biological engineered threats like pandemics have increased dramatically. That question, I will believe there is a high risk of AI takeover.